Jamie Chen
Rice University
Houston (29.7° N, 95.3° W)

featuring Carrie Ott-Holland, Manager, Performance Design Research, Google, San Francisco (37.7° N, 122.4° W)

Carrie Ott-Holland is an industrial and organizational (I/O) psychologist who manages a team of researchers redesigning performance management and promotion processes at Google in San Francisco. She graduated from Rice University in 2009 with a bachelor’s degree in psychology and music. Between 2010 and 2015, she earned a master’s and a doctorate from Michigan State University at the nation’s top I/O psychology graduate program. I had the pleasure of interviewing Dr. Ott-Holland in San Francisco through a video conference while I was across the country in Houston. In the interview, we talked about her unique career path and the two “existential crises” that led her to her current work at Google.

Coming from a preprofessional fine arts boarding high school, Dr. Ott-Holland’s precollege years were focused intensely on classical music and vocal performance. Dr. Ott-Holland came into college as a voice performance major in the prestigious Shepherd School of Music, but soon found herself and her goals to be quite different from her music school peers. Two years into college, when she realized that she did not want to spend every moment of her day studying music, nor did she wish to become a professional opera singer, she had her first “existential crisis.” She realized that she wanted a more diverse college experience with a major that would open her up to a wider array of options after graduation.

With the decision to leave music as a profession, she started to explore various areas of psychology by working as a research assistant in several on-campus research labs. During her senior year, through one of her many research assistantships, she decided that she really enjoyed I/O psychology and wished to pursue a career in the field after graduation. However, she realized that she didn’t have enough time to submit applications to graduate school for the coming fall. This “late” realization, however, became her second “existential crisis.”

When she graduated in 2009, Dr. Ott-Holland decided to spend the next year applying to graduate school while working part-time and taking graduate classes at Rice. However, the economy was in a recession, and Dr. Ott-Holland spent over six months searching for a job. Since she had graduated with a double major from a nationally ranked top 25 university, she had not expected rejection letters from everywhere she applied (including coffee shops). In the meantime, she started cold-calling members of the Houston-area I/O psychology professional association, asking for informational interviews to learn about members’ jobs, their graduate training, and professional advice for young professionals. One I/O psychologist encouraged her to apply for an internship at their company. This exchange resulted in an internship at a personality assessment company, Birkman International. The experience allowed her to gain crucial initial industry experience in the field of organizational psychology.

That year, Dr. Ott-Holland applied to over a dozen I/O psychology graduate programs. She was unsure of her competitiveness as a candidate and wondered whether she would be accepted to any program. To her surprise, she was accepted into Michigan State University, a top I/O psychology program in the country, which served as another crucial stepping-stone in her career.

During graduate school, Dr. Ott-Holland applied and received an offer to do a PhD internship at Google in their People Operations department, which eventually led to a full-time job after her graduation. Dr. Ott-Holland described Google as a company that values data-driven insights and holds a positive attitude towards research. She strongly emphasized the importance of Google’s “science-based HR” mentality for her, as it helped her bridge the gap between her academic training and the realities of corporate life, where purely profit-based business mindsets often drive decision-making.

When asked about advice for students in college without a clear goal or path in mind, Dr. Ott-Holland described a concept borrowed from algorithms that applies to careers—exploration versus exploitation. Although exploring different options is extremely valuable, at some point, when you find something that is good enough, you should just go for it. Both “modes” have their pros and cons. With 100 percent “exploitation” mode, you may reach your goal faster, but you run the risk of realizing your end goal may not be where you want to be, after you have already invested greatly. With 100 percent “exploration” mode, however, you may never decide a path to take and hence never achieve anything. I think this is a practical dilemma (and fear) many young people face today. While there is no clear answer or general rule of thumb, Dr. Ott-Holland’s journey shows that, as undergraduates in college, we still have time to explore and find something we feel interested enough in to want to pursue a career in it. And once you find something you are comfortable with, which balances both practicality and personal interests, you can move along with purpose, and things will slowly fall into place.

Dr. Ott-Holland also had some advice for students on the stress of finding an internship. She recalled, “It’s good to apply to internships, but don’t put too much pressure on yourself early in your career to be remarkable or secure the perfect opportunity. Life is long! Enjoy the opportunity to explore and reflect. The summer after my sophomore year, I worked as an office admin in Rice Village, read books, journaled, went for long runs in my neighborhood, and spent time with friends. Not saying it’s the best thing, but there’s nothing wrong with it.”

Douglas Graham
Rice University
Houston (29.7° N, 95.3° W)

featuring Gabrien Clark, Analytics Lead, Nike, Beaverton, OR (45.4° N, 122.8° W)

Gabrien Clark is an analytics lead at Nike, where he and his team work with sales data to find performance trends in the marketplace and communicate their findings back to the company. He graduated from Rice University with a bachelor of science in bioengineering and a bachelor of arts in history. During college, Clark was active in several student groups, including the Rice Student Association and Rice Empower, a science outreach program. Clark joined General Motors as a business intelligence developer after he graduated, and he tackled complex problems in a fast-paced working environment before joining his analytics team at Nike in 2018.

We began our meeting with a conversation about Clark’s path from studying bioengineering at Rice to becoming a data analyst. It started with a junior year technology internship, which opened his eyes to how his bioengineering problem- solving skills could be applied to different fields. Clark subsequently joined General Motors after graduation and was immediately assigned important projects that put his skills to the test. He went on to work at Nike, a company that shared his appreciation for creatively using different software to solve problems in favor of being constrained to any specific tools.

After Clark discussed his road to Nike, he explained how his current team works together to reveal insights about company sales and present their results. They first gather and interpret data from Nike, along with a variety of retailers, and proceed to determine patterns about product sales in the marketplace. But discovering these trends is only one piece of the puzzle: Clark and his teammates must also visualize and effectively explain their findings to the company and its stakeholders. Communicating these important findings can significantly affect Nike’s planning process. On data analytics teams like Clark’s, there is a high demand for data analysts who can confidently use data analysis and visualization software and work effectively in teams. Clark advises that students interested in the field should consider building a portfolio of projects using a variety of programming languages or software that they can show to potential employers in their job search.

An interesting takeaway from our interview was that the problem-solving skills you learn through one college major can be transferred to other fields in surprising ways. For instance, Clark’s experiences on bioengineering teams allowed him to quickly adapt to team-based projects at General Motors. Though his switch in fields demanded that he quickly learn the database query and management language SQL, Clark’s broad problem-solving skill set that he fostered at Rice enabled him to conquer the challenge. His experience showed me that although some career path switches may be more difficult than others, the hard work required in any rigorous course of college study gives students the ability to transfer their skills to a large range of industries.

But perhaps the most important lesson I gleaned from our conversation was that unexpectedly tough situations often lead to unexpected growth in skill. When Clark obtained his first job after college at General Motors, the company was rapidly shifting their team responsibilities in the wake of their 2014 recalls. Because of this, Clark was assigned difficult cost negotiation tasks that were pivotal to multimillion-dollar company operations. With the help of a guiding mentor, Clark applied his engineering discipline to the work and solved complex tasks, honing his data analytics skills along the way. Facing and overcoming such a significant challenge gave Clark the skill set he needed to succeed in his current position at Nike. His daily work includes adapting to unique situations and using new data technology, abilities that he gained through hard work and determination. Clark’s road to Nike demonstrates that with the right mindset, students can transform their career paths and emerge as skilled professionals in innovative industries.

Highlights from the interview:

Everyone’s family, community, and life circumstances create an initial role for them in society. What was expected of you, and did you stick to that path or stray from it?

I guess I was lucky. I’ve never really had specific expectations for where I would end up, career-wise. I was lucky enough to have a family that was very supportive. Super early on, in high school/middle school, I wanted to be a fitness businessperson. After a while, I wanted to go into medicine, and after that, I was kind of lost, didn’t really know. Family support really helped out, and eventually when I started getting on a path to technology and data analytics, everyone was super supportive.

At which point during your educational or professional journey did you begin to envision yourself in data analytics?

I took a tech internship during the summer after my junior year of Rice. That summer was one of the big influences, and during my senior year, I was pretty open-minded about what I could possibly do. Luckily enough, we had some pretty great technology partners come to Rice to do on-campus recruiting. One that really caught my eye, because I studied bioengineering at Rice, was someone from General Motors who was previously a civil engineering master’s from Rice. He talked about his transition from engineering to tech and mentioned how it is essentially all about problem-solving, even with technology. It got me very excited, knowing that I could take my problem-solving knowledge from my engineering education and bring it to just about any other field. Data analytics within technology has tons of complex problems that people were trying to solve, so it felt like a really great fit, where I could take—not necessarily my exact education—but a skill set that I honed [during] my education. I could take that skill set and bring it over to this new career path. That was a really big, important moment for me.

Was there any specific project that you worked on in data analytics where you really felt skills being transferred from the engineering classes you had?

I would say literally my first project in my first job at General Motors. General Motors had some pretty big vehicle safety issues at the time I started. So my analytics team there picked up newly added responsibilities. All the experienced people were placed on the priority for the vehicle’s safety, but all the existing projects that they still had on their plate had to go somewhere. Because of that need, my first project was actually pretty complex. It required information and tools to negotiate parts from our suppliers and to determine what the ultimate cost of a part was, including transportation cost. How much did that cost? What is it costing the company? And helping the manufacturing plants negotiate. It was actually a partnership with the research and development team. It was very difficult, but it was one of those things where you have to say, “Okay, these are the constraints. These are the technology I’m using. This is the current data structure. These are the things I can do with that data structure. Here’s ultimately what a report can look like.” And once getting those constraints down, going through the iterative design process with partners in research and development, testing out different prototypes for designs and basically getting to final product. Then working on any improvements, clearing the bugs and maintaining that product, and eventually ending up at what we see in production. That was, I would say, pretty much from the get-go, transferring those skills, applying those rational design practices I learned in my engineering discipline at Rice, and really iterating and building something really cool that’s useful for business. 

When you were solving all those complicated problems, did you have a particular mentor or person who helped you along the way?

Yes, I had a few. Luckily, my hiring manager helped me get set up. So, when I first started, I had someone who had been there only for a year but had built a really nice orientation for doing things the right way, interacting well between stakeholders and managers in meetings. So he really took me under his wing and really helped me out along the way. One of the other groups of people that I really appreciated was this group of people called the database administrators. They’re people that are basically the godfathers of the data, where we put it and share it. They’re really cool people; they’re all kind of older and super nice. When I started my job, I didn’t know SQL. I didn’t know how to design my queries as well. They did a good job of being very patient and helping explain things, help me learn along the way. It definitely made me a lot stronger than I otherwise would have been within the technology field. I’m very grateful for that because I know a lot of places don’t have those people that are willing to help you out like they helped me out. That’s definitely something that’s an advantage in my career, and a big part of the reason why I got my job at Nike, because I knew a lot more than the typical two- to three-year person in terms of optimization. I was very lucky that I was able to find those groups of people to be another set of mentors for me.

What lead you to your current position at Nike?

I had worked at General Motors IT for two and a half years. I went from level zero, not knowing anything about that field, and not knowing the different skill sets that I would need to be successful when I first started, to the point where not only did I know now, but I was hungry for more. I wanted to use more than what was available to me. I started learning Python, I started learning more advanced SQL, how to build more on the side of open-source things at home. And I was using websites like Codecademy and DataCamp to level up. Then, I was [contacted] by a staffing company, which had a contact with Nike. The team here at Nike was basically the dream. What he was explaining to me was they didn’t believe in using any predefined, set list of tools. They believed in using whatever gets the job done the best way. Not being afraid to explore, which is something that I really, really wanted. On top of that, just from a life-experience standpoint, I’d grown up in Texas, born and raised, went to school there, had my first job, all within Texas. So, I thought, you know, Portland would be a cool place to figure out, get some new life experiences. So, it was very fortunate that the right opportunity found its way to me via LinkedIn. In the technology field, people reach out all the time, but it was kind of one of those things were nothing seemed right until that moment. And it just worked out. I’m very fortunate that the people reaching out had a great opportunity for me.

What does your current position entail on a daily basis?

Well, it varies. A lot of what my team does is work with data from our retailers that Nike works with, as well as our own internal direct data. My team in particular focuses primarily on point of sales data. For example, we’ll look at a Nike shoe, we track it, and say how much has this shoe sold through its inventory across all of these retailers. And we’re able to compare it and know the key performance indicators (KPIs) to give us insight into how exactly we did in terms of planning for that shoe. Also, examine things from an inventory standpoint and look into what we could do to improve, based on what we’re seeing in the trends across our retailers.

How do you work together with people on your team?

We work in two-week sprints, where we have a product manager who determines a set of smaller items that we’ll be working on to improve our product or products. Our team will take these individual tasks and work on it, and complete them, and move them over into to-do or in-progress or done stage. That’s a very simplified version of it. During the workflow, we talk to each other and help each other out. We communicate whenever a new test comes up. We sit together and we estimate how much work we can get done. Over time, these meetings are where we actually do estimation and talking through the tasks. We get to a point where we understand how much work goes into the product creation, what tasks we can share, what tasks are better worked on individually. Whenever we are working together, we just hop over to each other’s desks, bring our computers, have any discussion we need to talk about for the particular work at hand. And then when we pass teams along, we take notes, detailed notes, built off of the things we had been doing in the past. 

What skills would you say you use most commonly in your work?

From a technical, hard-skills standpoint, I would say SQL and Python are very important for me. We do most of our data visualization in Tableau. Also, there’s a tool called Microsoft SQL Server Analysis Services. I know that may not mean much, but we use that quite a bit here and in my previous job to build something that’s called ad hoc reporting. So, a large set of people may have a shared question that could be answered with one set of visualization. Ad hoc tools allow you to basically just play with the data and answer your own questions.

From a soft-skills standpoint, I would say communication is huge. It’s very important that we’re able to talk to each other and discuss issues and be able to quickly resolve them. I would say communication within our team is very important.

Another important thing is something called business acumen. One important thing that I had to learn is that learning SQL, Python, all these other languages, JavaScript, whatever it may be, is extremely important to be able to create your products, but at the same time, the other big half of that is you really have to build up your knowledge of the data that you’re working with. In terms of Nike, that’s like really understanding how Nike conducts its business, what we deem valuable from a metric standpoint in terms of point of sales, or finance, or what we stand for in a total marketplace. So understanding those key metrics, how our business analysts look at them, and how we’re using that to drive our company forward. Understanding the whole company’s supply chain process is extremely important, because when you’re meeting with stakeholders, you’re not meeting with people who are extremely technical, so you’re not going to be talking about the ins and outs of your SQL query. What you’re going to be talking about is logic. That’s the common language. Someone who’s working within the business and making business decisions will have a logic in mind, where we look at this metric like this and this is how we feel like calculating it. To yourself, you have to translate that to a query or a calculation, but when you’re talking to a business stakeholder, you really need to be able to talk at the lowest level, the logic standpoint, and realize why it’s important and how it’s used.

Were there any skills from your college years at Rice that helped you in those conversations?

I think in engineering, one of the important things that we need to do is make things concise and succinct and to the point—and back it up with numbers and tell a very good, clean story. In terms of working with business stakeholders and keeping our communication of our team to the point and effective, I really need to use those skills I learned in college. I think it’s important for any other job as well, but yeah, being able to concisely, succinctly, and very effectively tell your story… 

How are science and technology reshaping the work that you do, and what changes do you foresee in your industry coming up? 

I think probably the biggest one is something that’s been coming for a long time, it’s constantly building, and here it’s all about cloud computing. It’s one of those things where more and more companies are doing a great job helping offload the computational needs of the company. They provide these great services that scale up very nicely and are extremely important. It’ll blow your mind if I were to show you, just from a reporting standpoint, how much stronger we got by moving to a cloud-based computing platform.

I think also what comes along with that is this idea of software as a service is also getting really big, that model where everyone is trying to be like a Netflix or Spotify, where they offer you monthly services or whatever time period. They offer you these services, but you pay a monthly fee. You scale up or down what you pay for, and your services go up and down. I think those two things go hand in hand, and I think that’s something that the field is just generally moving toward.

I think one thing that won’t be changing anytime soon is people being the backbone for data and analytics. I think it is extremely important. That’s the centerpiece of everything for me. That’s one of the most important things that anyone going into this field can learn. SQL and knowing it well, understanding your data structure well, and if you’re using cloud-based computing platforms, really understanding the ways that they actually compute and what ways can you improve the performance of what you’re building. With those platforms, understanding how much it’s costing us and in what ways we can reduce those costs—and, with that, still really great products.

What would you say is the biggest challenge facing your industry?

I would say it’s a nice challenge, but I still think that from a demand-supply standpoint, that the supply still isn’t meeting demand in terms of talented workers who come in and do a good job and help companies get along. I think being able to really get what teams need from a human-resource standpoint is going to be a big challenge. I think that’s one of the nice challenges; it’s not a typical challenge. It’s nice for workers, but it is a real challenge companies to face. One of the bigger problems is when you have some big initiative, then you have to hire for that, then you have these gaps where you have to go through a full process. For a big company like Nike, that could take quite a while, so these can often get sidetracked a bit. I think that’s honestly going to be one of the bigger challenges.

What advice would you give to a student interested in your field?

If you want to go into data and analytics, learning SQL, building products on your own, or building projects with a team, and building a portfolio that you can walk through and speak through and give details on how you contributed to building that project out. I think that’s hugely important because, ultimately, at the end of the day, someone who’s hiring you, the thing that they want to know is that you’re going to be able to take the job, and do a good job, and very effectively do everything that you say that you can do on your resume, and also things that they are looking for that you don’t know. Be able to pick up on [new things], be willing to learn, and educate yourself. And I think by building up your own data analytics-focused project, as well as building up that portfolio, would be a huge helping hand for getting hired. I think that’s one of the things that every hiring manager will have to see, a portfolio and a project that you’ve done.

Interview excerpts have been lightly edited for clarity and readability and approved by the interviewee.

Douglas Graham
Rice University
Houston (29.7° N, 95.3° W)

featuring Dr. Karen Book, Senior Data Analyst, Boston Public Schools, Boston (42.3° N, 71.0° W)

Dr. Karen Book is a senior data analyst at Boston Public Schools. She works on the planning and analysis team, where she evaluates and presents data to support the school district’s enrollment planning. She graduated from Rice University with a sociology degree in 2011 and earned a PhD in sociology from the University of North Carolina at Chapel Hill in 2016. While earning her doctorate, Dr. Book worked as an associate editor at the journal Social Forces, as a predoctoral trainee at the Carolina Population Center, and as a teaching fellow at the University of North Carolina at Chapel Hill.

Our interview began with a conversation about Dr. Book’s path from college to her current profession as a senior data analyst. At Rice, she discovered an interest in understanding inequality with data investigation and for that reason decided to pursue a PhD in sociology. While she continued to enjoy analyzing data to study education and inequality in her doctoral program, Dr. Book found that the emphasis on publishing in academia did not align with her values. Hoping to take part in actionable research with direct impacts on communities, she became a data analyst, which allowed her to use the skills she learned in education, such as Stata and Microsoft Excel, to tell stories with data. By using data to show trends in school enrollment, Dr. Book and her team contribute to district decisions that result in positive outcomes for their community and school district. Dr. Book emphasized that software programs allow analysts to process data more efficiently and make more sophisticated diagrams and conclusions.

One significant piece of information I took from our conversation was that if you have an interest in a field of social science like sociology, learning data analysis through software programs will open interesting doors for your career. Dr. Book explained that data is the underlying language being used in many social science fields; knowing how to use it effectively is key to understanding and communicating ideas within those fields. In Dr. Book’s career, her unique combination of expertise with the technical software Stata and her passion for combating inequality in communities set her on the path to a profession she finds meaningful. I found this interesting because social sciences are often separated from STEM fields in college, but there are many jobs where proficiency in both is necessary. I also learned from Dr. Book that it is perfectly fine, and sometimes necessary, to stray from the specific academic path you have set for yourself as you discover more about what kind of work you find rewarding. Dr. Book’s decision to pursue data analysis in education instead of continuing in academia was difficult for her, since it involved changing her career direction and having tough conversations with her colleagues, but it was key to finding a career that corresponded with her values.

In addition to our discussion about career paths, I found Dr. Book’s current projects at Boston Public Schools and how she uses certain skills from college and graduate school in her work particularly compelling. She and her team pull, organize, and analyze data on student enrollment in particular schools, programs, and grades. Then they find patterns in data so the district can decide whether more schools or classrooms are needed for a particular neighborhood or program. Throughout the process, data serves as the foundation for important planning conversations at Boston Public Schools. Dr. Book uses this data to communicate a story that engages community members and school district officials. Her work results in outcomes that benefit students, families, and educators. From Dr. Book’s description of her work, I gained an appreciation of data’s importance in supporting education and uplifting communities. All in all, our interview taught me that there are multiple ways you can explore your field of interest, both inside and outside of academia, and the specific path you should choose depends on what you value and find rewarding.

Highlights from the interview:

Everyone’s family, community, and life circumstances create an initial role for them in society. What was expected of you? Did you adhere to it or stray from it?

I think my family always had high expectations for education. I was always a good student. I enjoyed school, and that led well through high school and into college. When I was in college at Rice University, I discovered that you could make a whole career out of being a researcher, and that’s when I decided to go pursue my PhD in sociology at the University of North Carolina (UNC).

While I was at UNC, I thought I was on that path to be a professor, to be an academic researcher, and things were going well. I was again pretty good at it. You say you have certain paths set for you…I got onto that academic path early in graduate school, but it was probably my third or fourth year of my PhD program that I realized that while I really enjoyed sociological research and data-driven study of education, inequality, life course studies, and that kind of thing…I was finding that the value structure and the reward structure in academia was something that I was not enjoying, and that’s the stereotypical “publish or perish.” There’s a huge emphasis on publishing, and I interpreted that a lot of times as publishing for the sake of publishing, and publishing for the prestige of publishing, rather than publishing data or research that could be shared and built on. I much more enjoyed that part of it versus the actual following through and publishing of things. That’s when I had to break from my path.

I had an incredibly supportive mentor throughout my graduate school years, and when I told her I wasn’t sure If I wanted to follow in her footsteps and become a tenured professor, tenured research professor, that was a hard conversation to do, but I’m really glad I did it, because the work I’m doing now much more closely aligns with things I value in terms of what’s rewarded, what’s valued, that kind of thing. And doing research that gets used.

When did you start to envision yourself as a data analyst for your profession? Was there a certain turning point?

I had always been a data analyst in a sense, and that was the part I really enjoyed. I loved our weekly check-ins with both my mentor in undergrad as well as my advisor in grad school. We’d talk about the data; I’d show them the latest regressions I’d run. We would talk through what does this mean, what does this mean. And there was always this sense—I want to build on it, I want to build on it. But the next step would be like okay, well, time to write it into a paper. I didn’t mind the writing of the paper, that was also an interesting exercise, it was having to wait six to 24 months…sending it off to different people. I wanted that instant gratification, almost. Things take time. At least in my role now, I think things are taking time because they’re still being built on in an actionable way versus just a paper floating out, sitting on some other professor’s desk or researcher’s desk, waiting to be reviewed and built on. The process was just too slow for me, and the outcome didn’t seem to be worth all that time for me.

What led you to your current position, and what do you do in your current position?

I was searching for jobs in Boston. I was doing an internet search of one of those—Indeed or LinkedIn—where you can type in keywords and job postings will come up. I typed in the software that I use, which is Stata. It is used a lot in economics and in sociology, as well as other academic fields. It’s statistical software that is akin to R or SAS or these softwares that are more well-known, especially in the nonprofit, government, private sector. I searched for Stata, because that’s what I was good at, that’s what I know how to use. And a job at Boston Public Schools popped up, and I clicked on it. As I’m reading the job description, I’m like, “Oh my gosh, this job is perfect for me.” It was using data to explain things, to help identify inequities, et cetera. The job has shifted since the original job description, which is why I’m being a little vague. I saw this and thought, wow, they value someone with a strong quantitative background, but it’s also an organization that’s doing important work every single day and is directly impacting the way we’re serving the students and children of Boston, and so I applied.

What are some of those improvements that you mention your work is doing?

My current role now is I’m on the planning and analysis team at BPS. In summary, the work I do is using data to support short-term and long-term planning in the district around enrollment. We use enrollment data, what students are attending what schools, in certain programs in various grades. We look for patterns in those ways. We provide data on what programs are growing, and we might need to open new classrooms. We also identify what parts of the city have fewer seats relative to the number of students who live there—so there’s a capacity shortage, so that’s where we’d recommend building new schools or opening new classrooms in these areas. That’s just one of the many things our team is doing as part of a larger process. I’m not the one making those decisions, not by far, but I’m pulling the data, and I’m organizing the analyses we’re doing.

Other than your knowledge of Stata, what are some other skills that you find yourself utilizing a lot in your position? How did your college years prepare you?

I think the skills that I learned in college and grad school, as well for a little bit [on the job is]—how to tell a story with data. I think, at the end of the day, that’s what appealed to me about being a professional sociologist or a researcher. And then what is further parlayed into this career being so rewarding for me is interpreting numbers and making it accessible to an audience, whether that’s through writing a memo or a paper as I may have done in college or a presentation that I did in college. Now I’m doing a similar thing. I’m taking data, and I’m helping people understand what the data means. I’m making it accessible to a layperson or somebody who maybe understands a lot more about what’s happening inside of the schools. I can bring them information about enrollment patterns that maybe they had seen to some extent but didn’t really fully understand until I put it in certain ways. I walk people through it, answer their questions. That’s something that’s always really appealed to me.

Are there ever any misconceptions about your profession or about what you do?

I think a lot of people might hear “analyst” and think that we’re doing really high-level, super in-depth statistical analyses all the time and that we’re just in front of our computers all day, not talking to anybody. And there are definitely moments when that’s happening, and I’ll have my scripts running, and they’re super intense, and people are like, “What is that? What are you doing? That’s crazy.” But a lot of what I do is a lot more simple descriptive statistics. I think the power of descriptive statistics and just simple charts can really be powerful. And, as I mentioned earlier, that telling of the story—sometimes the more complicated statistical model that you run isn’t necessarily the most important data that you analyze nor is it necessarily better because it’s more complicated. There’s just an element of analysts being engaged with data and doing analytics, but there is that element of translating it. How do we make it approachable to an audience? And that audience includes people that I work with in the central office, it includes principals, it includes school communities and families…Basically, there’s a lot more to being an analyst, at least in my role, than just being in the data all day long and doing crazy complicated things.

What parts of Boston Public Schools make it a good place to work?

I am constantly in awe of how many smart and capable people I work with. On my own team, I have so much respect for the people I work with, and that makes it fun, it makes it good, makes it a reason to come into work every day. I really respect and value everyone on my team, but then that extends to other people in the central office. There’s really the sense of doing it for the students. The work we’re doing is grounded in schools, and the principals that I get to work with are just some of the most excellent people because they are engaged in making sure their schools are welcoming and rigorous and a place where children can feel safe and learn. So I think being a part of an organization like Boston Public Schools is great because I get to do the work that I enjoy. I’m doing it for an organization that I believe in, in the sense that everyone’s in it for the kids, and I think they’re doing a good job. I mean there’s always room for improvement. BPS, like all urban school districts, faces so many challenges, but I really respect the people I work with. I think that makes it a great place to work.

How are science and technology reshaping your work, and what changes do you foresee in your specific area?

Let’s just say that a lot of the statistical power and data that we have now, we didn’t have 10 years ago. I wasn’t here 10 years ago, but I hear stories of how things were 10 years ago. There was that same, “Oh, we want to be equitable and do things in the best way we can,” but there wasn’t that level of data infrastructure and people around who could interact with it in the same way. So I think the fact that we can run scripts in R and Stata, generate a bunch of tools…that kind of thing wasn’t being done in the district five, 10 years ago. Having the additional machine power behind us has really enabled us to tell the story faster, with more intricate charts and reports. I think our leadership really values data as something that’s really important and valuable, and that’s part of the reason our team has come together and has thrived. We have all this data, we know how to use it, and people in the district value it and value the work we’re doing.

What advice would you give students interested in the field?

I’ll give some broad advice, which is if you have an interest in analysis but you also have an interest in urban education or sociology or insert whatever non-STEM field, I think as data is becoming more and more prevalent, it’s the language that everyone’s using in almost all fields these days. I didn’t even know a job like mine existed when I was in college. And maybe it didn’t. Maybe it’s a newer job. There has been this increase in jobs that use data as the quantity of data has increased. But I think I would challenge or advise students who are interested in this kind of work to also think about where your passions are—I know that’s kind of cliché—because there’s just data and analyses happening in every field. And so, not limiting yourself to working at a tech company or for a consulting firm or something. There are people doing that work in every field, and I think pushing people to think through, “How could I be doing data work in a field that interests me? Or a field that I feel passionate about?” like urban education or whatever thing that is interesting to you. Those jobs are out there, and the numbers are growing, so I encourage students to keep up with the latest trends in data and analyses and the latest software. Data visualization is becoming hugely important, so those kind of skills—brush up on those skills. Think about the industries you might not think about as having roles for data analysts. I bet that they do.

I would want to give students permission to stray from the path that they think that they’re on. If you think your major is setting you up for only a certain type of job, or you think that your interests are only setting you up for a certain type of job, don’t be afraid to think outside the box. Network with people and things that you didn’t think you could do with your major. Spend a lot of time thinking about what you want to do, and don’t just stay on the path that someone is laid out for you because you have certain skills or you’ve taken certain classes so far. Obviously, your major and your classes are important, but the longer you’re in a job, the less relevant they become. So if you can establish that you have the skills and the interest in jobs that maybe you didn’t think about before, you’re qualified for more jobs than you probably think. If you spend some time thinking about that and thinking off the path, I think there’s a lot of potential for really cool opportunities out there.

Interview excerpts have been lightly edited for clarity and readability and approved by the interviewee.

by Steven Feng, economics and global health technologies sophomore at Rice University, Houston (29.7° N, 95.3° W)

The buzzword “data” has recently become commonplace in the academic, professional, and everyday world, and discussions on data and its applications are ongoing and striking. In the current age of technology, where everything is integrated with some sort of hardware or computer program, avoiding data generation and collection is impossible. Every time a student receives an email, a parent buys a toy for their child on Amazon, or a researcher narrows their search criteria on an academic database, data is being generated and collected.

Though this data is inherently meaningless, professionals and academics can use computational and statistical techniques to analyze and create meaningful and significant conclusions from data; the study and practice of these techniques is known as data science. In the words of Pierre Elias, a cardiology fellow at Columbia University who also conducts research in data science, data science is the combination of three disciplines—computer science, statistics, and content expertise (the particular subject matter a specific data scientist is interested or knowledgeable in). According to Elias, successful data scientists are not necessarily experts in all three fields; most have a unique balance between the three which allows them to collaborate successfully with other data scientists who have slightly different skill sets.

Currently, professionals in nearly every field are integrating data science into their practice to optimize or modify current methods. Two commonly cited examples of data science are within marketing and transportation. Companies like Facebook and Google are notorious for compiling and analyzing user data to personalize advertisements for individual users and oftentimes are under scrutiny for the sly ways they collect and access this data. On the other hand, transportation companies such as Uber and Tesla use spatial and geographical data to reroute drivers based on lower driving times or mileage or to power self-driving cars.

Many other fields have been using data science for their benefit for a long period of time yet do not receive as much media attention as the others. One such field is health informatics, the collection and study of patient and clinical data in healthcare. The healthcare world spans multiple subdisciplines, including pharmacy, private healthcare institutions, academic institutions, and insurance, and within these disciplines are great amounts of data. There are substantial problems that can be improved on by the analysis of said data using data science techniques. Integrating technology into the long-standing world of healthcare has been a work in progress, and recent developments in the field of health informatics prove how much untapped potential still lies in the exploitation of health data.

Professionals in health informatics are primarily concerned with optimizing or overhauling existing healthcare infrastructure because errors in healthcare practices and physical capital (that is, machines and hardware used in healthcare, ranging from fax machines to MRI scanners) that create inefficiencies can be addressed with or remedied by health informatic techniques. Both Elias and Julian Yao, senior director of strategic initiatives at Covera Health, a startup that uses data science to improve patient care, expressed dissatisfaction toward the current state of physical infrastructure in healthcare. Yao likened the industry as a whole to an operation stuck in the 1970s, and Elias commented on this issue further by mentioning how he still receives large data sets through a fax machine and sometimes has to comb through the documents manually in order to extract relevant data. In this example, data science techniques can completely obsolete the existing capital by providing a streamlined method for practitioners to compile, organize, and send data electronically.

Another important source of inefficiency in healthcare is misdiagnosis, which may be avoided with more advanced data science practices. When doctors diagnose a patient’s symptoms, their reasoning can possibly be based on false or misleading data points; according to Elias, most patients do not describe their symptoms in enough detail, which can make a diagnosis less accurate. Even with adequate patient description, however, misdiagnosis can still happen: advanced images such as MRIs are often inherently difficult to interpret, and human error by doctors is always possible. Covera Health studied the differences in diagnoses from different practitioners by sending one patient with lower back pain to 10 professionals in the greater New York area. Shockingly, not one diagnosis appeared on all 10 reports, and according to Yao, “if you take the two most extreme reports and put them side-by-side, they don’t even look like the same patient.” A misdiagnosis can cost a patient valuable time and money, and it can lead to further medical complications if the patient undergoes treatment for a condition they do not have. It can also damage a doctor’s reputation and place them under considerable legal and financial pressure.

Though healthcare informatics is not the be-all-end-all, proper data science and machine learning techniques can significantly alleviate these problems. Computers, for example, can be trained to scrape, or extract, data from files. To do this, a data scientist would first design an algorithm telling a machine what sort of information to look for and then train the machine by feeding it data and, to put it simply, tell it what is right and wrong. For Elias, a machine that automatically compiles patient data is a considerable upgrade from his method of receiving faxes and then extracting data by hand.

Machines that can diagnose illnesses are a trending research interest amongst practitioners and data scientists alike. Following the same machine learning principles described above, data scientists can train machines to analyze MRIs and other images for symptoms by “feeding” the machine examples of MRIs where symptoms are present or absent. With enough data points, the machine can train itself to detect symptoms from new images. This automatic process can greatly improve the issue of misdiagnosis if the technology is trained properly and thus is able to detect conditions with any given MRI. There are constantly new developments with machines that can assist practitioners in diagnosing conditions. At Covera Health, for example, Yao and his team specifically tackle misdiagnosis in radiology by amassing clinical data and then analyzing it not only to improve diagnostic accuracy but also to ensure patients get the optimal care in order to improve outcomes. In addition, Elias mentioned developments in machine sensors to better interpret images from echocardiograms, and researchers at Stanford have developed an algorithm known as HeadXNet to detect brain aneurysms through MRIs. Groups of data scientists leverage that same core trifecta of computer science, statistics, and content expertise to effect life-saving changes in an industry long due for a technical overhaul.

New developments in healthcare informatics will take some time, given how arduous and time-consuming the process to gather data, develop algorithms, and train machines is. In the meantime, both Yao and Elias offered a common piece of advice to undergraduate students: learn data science. As more tasks are automated, data science is becoming more and more relevant and intertwined into every professional field, and the value of a data science background cannot be understated. Even still, data science will never be a one-stop-shop to solve all of the world’s problems but rather an important method in doing so. Elias stressed how data science is not a silver bullet destined to fix everything. Strong and reliable data science applications are on the way, however, and developments are only getting better and better.

Further Reading:

• The key word in “Data Science” is not Data, it is Science – In a quick yet informational blog post, Jeff Leek provides a quirky perspective on data science while describing parts of the overarching process of gathering and analyzing data. https://simplystatistics.org/2013/12/12/the-key-word-in-data-science-is-not-data-it-is-science/
• Stanford researchers develop artificial intelligence tool to help detect brain aneurysms – Taylor Kubota describes the development and use of the HeadXNet algorithm by medical professionals.
  https://news.stanford.edu/2019/06/07/ai-tool-helps-radiologists-detect-brain-aneurysms/
• Variability in diagnostic error rates of 10 MRI centers performing lumbar spine MRI examinations on the same patient within a 3-week period – The clinical study from which Yao draws his lower back pain anecdote. https://www.thespinejournalonline.com/article/S1529-9430(16)31093-2/pdf
• How Covera Health Is Redefining Quality Standards in Radiology – In this article, Kenny Kline, Founder of JAKK Solutions, describes Covera Health’s operations in more detail and describes their partnership with Walmart.
  https://www.inc.com/kenny-kline/how-covera-health-is-redefining-quality-standards-in-radiology.html
• Clinical research informatics: a conceptual perspective – This journal article defines and visualizes the process of clinical research informatics, referred to in this work as health informatics, starting with posing questions from “transitional discoveries” and then following through by collecting data, formulating hypotheses, and gathering and implementing evidence.
  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3392857/
  

Longitude.site welcomes applications from students who are interested to explore other topics related to data science and healthcare. Apply here.

Hallie Trial
Rice University
Houston (29.7° N, 95.3° W)

In the late 1800s, tissue-staining dyes began to captivate physician-scientist Paul Ehrlich. In particular, he noticed that some dyes stained only bacterial cells while leaving mammalian cells their natural color. Erhlich envisioned a toxic dye molecule that would bind to and kill bacteria but leave human cells perfectly healthy; such a molecule would cure bacterial illness. After testing more than 600 arsenic-based dye compounds, Ehrlich discovered the first modern antibiotic, arsphenamine. This compound allowed doctors to treat syphilis effectively for the first time. 

Just like Ehrlich’s dyes, all drugs operate by selective binding: they stick to some molecules but not others. Penicillin, for instance, attaches to and inactivates a vital protein that bacteria need to synthesize their cell walls, but the compound is harmless to human proteins. Aspirin helps control pain and swelling by binding to the enzyme that produces hormones called prostaglandins and thromboxanes, which serve roles in inflammation. 

With the ever-advancing modern understanding of biochemistry, scientists can pinpoint exact proteins involved in causing disease. Our current approach to drug development is to select a druggable target protein and then maximize the binding of a drug to that single protein while preventing other interactions. Ideally, minimizing effects on other proteins minimizes side effects. In the complex, intricately connected systems of living organisms, however, there are often multiple redundant pathways to achieve the same purpose. This means that when a drug alters the activity of just one biological molecule, other molecules can sometimes fill in, and the medication may not produce as strong an effect as expected. Furthermore, biological molecules influence one another’s activity in complicated ways, so unexpected side effects can emerge. 

A new approach to drug design called network pharmacology considers the interconnected living system as a whole rather than focusing on just one druggable protein. Dr. Anil Korkut from MD Anderson Bioinformatics and Computational Biology explained that network pharmacology “focuses on how pharmacological agents can alter the molecular network within a cell.” 

Molecules in a cell interact much like people within a large corporation. In order to determine which figures are the most important for the functioning of the corporation as a whole, one might consider who talks to whom, how their communication influences each person, and whether people could still get the information and orders they need without each communication exchange. Similarly, to investigate which molecules to target with drugs, molecular pharmacologists must use computational models to map the interactions between many different biological molecules and the pathways relevant to a disease. They might find one single, vital “CEO” molecule that they can target with a drug to alter the whole disease “corporation.” Other times, they might find a collection of a few important “employee” molecules that can alter the direction of the disease “corporation” when targeted together but not when working alone.

Scientists have investigated network pharmacology through experimental procedures like systematic screening and computational network analysis. In systematic screening, researchers study the relationship between only two drug targets at a time. For example, they might have one anticancer drug that they wish to combine with another drug—a co-drug—to improve its effectiveness. They could add the first anticancer agent to thousands of tiny cancer cell cultures, add a different second drug to each culture, and determine which two-drug combinations work most effectively. Alternatively, scientists can perform synthetic lethality chemical sensitization screening. This requires them to create many different cancer cell lines, each with one gene turned off. Then, they add their drug of interest to each cell line. Some cell lines will show increased sensitivity to the drug, and this tells investigators that the particular genes turned off in these sensitized cells might be useful targets for co-drugs. Screenings can often help uncover complicated, sometimes counterintuitive drug interactions and lead to improved medication combinations. 

Much of the future of network pharmacology lies in applying computational methods to biological networks and the chemicals that alter them. Advancements in computer models of the molecular networks within cells improve the efficiency of screenings by helping identify in advance which target combinations will likely effect the desired change. Once researchers decide which proteins or molecules to drug, cheminformatics databases correlating molecular structure with biological function and molecular binding can help chemists design compounds that affect those targets. 

An inherently interdisciplinary field like network pharmacology demands researchers from many different academic backgrounds. Projects need computationally-trained individuals with some knowledge of chemistry, biochemistry, or both, as well as experimental chemists and biologists with enough understanding of computational models to apply them. Most positions require a PhD in a related discipline, such as computational biology or chemistry, systems biology, statistics, biochemistry, chemistry, and many others. Undergraduate students can prepare for these diverse graduate programs with a wide array of majors, including bioengineering, biochemistry, chemistry, computational and applied mathematics, statistics, and even biological physics.

A few organizations currently specialize in network pharmacology. Aside from MD Anderson where Dr. Korkut runs his lab, he has highlighted groups at Harvard Medical School and Oregon Health & Science University. Harvard has a specialized Laboratory of Systems Pharmacology, and OHSU houses the Center for Spatial Systems Biomedicine, whose researchers conduct many projects in the related field of perturbation biology. As network pharmacology develops as a discipline, more opportunities may emerge both in industry and in academia. 

With further advancements in computational tools, network pharmacology may fundamentally alter how we think about chemical treatments for disease. It may also someday play a part in personalized medical care. If we develop the ability to create a distinct network model for each patient’s unique biology, Dr. Korkut stated, “we will be able to predict the immediate molecular changes in a given patient after therapy, and that will enable us to come up with better therapies, especially drug combinations… we are going to change people’s lives.”

Related Articles

This discusses the definition of network pharmacology, the goals of the field, relevant investigation platforms in the discipline, and future challenges. https://www.ncbi.nlm.nih.gov/pubmed/18936753.

This provides an introduction to networks in biology. https://www.ebi.ac.uk/training/online/course/network-analysis-protein-interaction-data-introduction/network-analysis-biology-0

The ACS page on cheminformatics provides information about the discipline cheminformatics, the questions it addresses, and careers available in the field. https://www.acs.org/content/acs/en/careers/college-to-career/chemistry-careers/cheminformatics.html

This research article authored by Anil Korkut and others provides an interesting example of network biology in action. The researchers constructed network models based on numerous experimental measurements to identify potential drug combinations for treating a drug-resistant melanoma cell line. https://elifesciences.org/articles/04640

Excerpts from a Longitude.site interview with Dr. Anil Korkut in student reflections.

Steven Feng
Rice University
Houston (29.7° N, 95.3° W)

featuring Julian Yao, Senior Director of Strategic Initiatives, Covera Health, New York  (40.7° N, 74.0° W)

Julian Yao is the senior director of strategic initiatives at Covera Health, a New York-based clinical analytics startup that uses advanced data science to develop meaningful quality metrics to improve diagnostic accuracy and help ensure patients get the right care from the start. Though his current career is now focused on innovation in healthcare, his time at Rice University as an undergraduate reflected various interests; Julian matriculated at Rice University as a music major at the Shepherd School of Music, but he eventually graduated with a degree in economics and international relations. At Rice, Julian was selected as a Baker Institute Public Policy summer fellow, participated on three alternative spring break trips, and served as the business manager for the Rice Thresher, the campus newspaper. After graduating, he joined Accenture as a strategy consultant before ultimately joining Covera Health.

Initially, my intention with the interview with Julian was to learn more about the intersection between healthcare and data science and how different companies like his are focused in this area; as an economics student myself, I was also curious about how Julian’s undergraduate lessons in economics and international relations led him to the career path he is on today. However, our conversation extended far beyond the intended scope, and we found ourselves talking about job searching and the importance of having a mentor. In short, what started out as an extremely technical and objective conversation eased back into a casual and cordial discussion about education, upbringing, and adulthood.

Prior to having a conversation with Julian, I did not fully understand the importance and applications of data science. Data science is a discipline that I always hear is “up and coming” or “important to learn” but never anything more specific. As I learned, companies in this space usually focus first on acquiring large amounts of data and then on employing different mathematical methods to analyze the data to uncover trends or patterns. The applications of data science are ubiquitous in our daily lives, from Google and Facebook collecting our data to personalize advertisements to self-driving cars learning the right and wrong ways to navigate the street. Julian and Covera Health use the clinical data they amass to create meaningful quality metrics to help improve diagnostic accuracy and ensure patients get the right care from the start of their care journeys. One of the key points Julian emphasized during our discussion was that undergraduates need to learn skills associated with data science. He noted that data science has been integrated into every industry and career path, and he regrets not taking more advanced statistics and computer science classes as an undergraduate. I mentioned Rice’s new data science minor to Julian during our time together, and he suggested that I—and other economics students—consider pursuing the minor.

Interestingly, talking to Julian about his time at Rice made me more confident about my future. While our conversation focused a lot on data science, a recurring theme in our conversation was the usefulness of “soft skills,” a term broadly describing interpersonal skills not acquired within the classroom. Drawing from his experiences at Accenture and Covera, Julian explained the times in his career where communicating effectively, being responsible, and leading groups of people were just as important as the computational techniques he learned in school. After a gruesome semester of calculus-based microeconomics and linear algebra, I felt reassured by Julian’s words that, while theorems and algorithms are important, they will not be everything in my career, and there is another skill set that I can continue to improve to complement whatever I learn in class. That is the beauty of a Longitude conversation: you can start a conversation at the intersection of healthcare and data science, but you can never predict what ideas and lessons you will learn along the way.

Highlights from the interview

You started your career path as a strategy consultant at Accenture, but you later moved into the startup side of the healthcare business, ending up now at Covera Health. During your upbringing before college, did you ever envision yourself working in healthcare, or working at a big corporation and a startup, and did your perceptions change once you had experience in college?

I actually went to Rice as music major—opera performance—my first year in college, at the Shepherd School of Music. Music is a big part of my life and coming into college, I wanted to explore if music performance was something that I want to really dive into. It was an amazing experience, but I realized very quickly that while I loved music, my passions from an academic perspective were outside of the performance halls. That was a huge turning point in my career.

I switched my second year into pursing a degree in economics and political science. Growing up abroad in Asia, I had always been interested in international relations and political science. I spent two summers interning in Washington D.C., one of which was through Rice University’s Baker Institute summer program where I ended up working at the thinktank Center for American Progress, working on international economic policy research. Going into those internships, I really bought into the idea that I could effect change in the world through policy. But what I quickly realized being in D.C. for two summers is that policy and politics are inherently inseparable and thus extremely slow-moving and cumbersome. That was another turning point in my career.

And literally on the second to last week in my internship before heading back to Rice, I ran into Kyle Clark on the street in D.C. Kyle, who I consider a dear friend and a mentor to this day, was Will Rice’s president when I was a freshman. Kyle was working at Accenture and urged me to explore consulting as a career. I was drawn to the fact that he, only a few years out of college, was effecting change through the private sector and building skills to bring value to his clients. I got back to Rice and started practicing case interviews, which are basically interviews that test how you think for these big consulting firms. I really enjoyed them because it makes you feel like you’re solving a meaningful business problem. I went through the interview process and accepted a role at Accenture. After I joined, I jumped at the first opportunity to be in the public sector and health services division at Accenture and have been in healthcare ever since.

Covera uses data analytics to prevent radiology misdiagnoses. Could you provide a brief example of a project you worked on that had to do with the data analytics side of healthcare? What skills did you and your team use to complete that task?

Yeah, but I think it’s good to put it in context. I was at Accenture for three years. I learned a lot about healthcare as an industry. I learned about who the main players are—provider groups, academic hospitals, insurance companies, different types of vendors—but also became an expert about a wave of new tech startups that many of our clients at large insurance companies were asking us about. At Accenture you get thrown in immediately—it’s pretty crazy—as a college graduate you’re in the same room, sometimes presenting to CEOs, CFOs, and big decision makers at large corporations. Anyway, many of them were asking us, “Oh, we heard about XYZ healthcare startup. What are they doing? What are they up to? How do we work with them? Do you think them as competitors?” It was really interesting for us, as consultants, to be in a position that monitors the field and to connect the lines to identify trends and which organizations were gaining traction.

After a few years at Accenture I really wanted to have the experience of building something from scratch, so I joined Spreemo Health, which is chapter one of the current company I work at. We had a growing operational business, but we were also concurrently building a robust data analytics infrastructure about measuring quality in radiology. I was brought in to tackle the question of, “How do we expand our business model to more clients in a different market segment?” Twelve months after I joined, we sold 90 percent of the company and spun out our analytics infrastructure and team into Covera Health.

We spun out [Covera] with around 10 people. Now 18 months later, we’re at over 40. We just launched a nationwide program for a Fortune 10 employer client and are looking to implement with new clients this year. We closed our Series A funding of $8.5 million. Looking back, the entire process, and project I was involved in, leverages something I learned from my college years

A lot of the skills around thinking about data analysis I’ve built and learned from my classes, internships, my experience as a management consultant, and I’m still very much learning now. I think the biggest lesson I learned is really about getting into the mindset of how do I step back and think of the data analysis as a whole and be able to specifically ask: what is the business question that we’re trying to answer here? Of course, at the very beginning of your career, it’s about how to create value as an entry-level analyst, right out of college. This will grow into thinking about data analysis from a high-level perspective and how to translate data analytics to make business decisions.

There are a lot of “hard skills” that you learn, but I’ve also learned a lot of “soft skills,” specifically around the art of sales through storytelling. For example, how do you sell a vision, how do you tell a story, how do you provide data behind that story to make it compelling, and how do you get the buy-in of not only the person you’re trying to sell but all of their vendors to make sure that your clients work with? A lot of these softer skills pertain to people communication. It’s really about how do you deal with people who have their own agendas, have their own things that they’re working on, and aren’t necessarily incentivized to work with you.

Can you describe the dynamics of a team that works on a project at Covera in terms of the structure, the organization? Are there important characteristics that make a good team in your company?

My answer is probably a little different from others because we’re a 40-person startup. I’ve been lucky in that I’ve been in a situation to hire people to build out our team, including people who are higher on the totem pole than me. As a founding member of this company, it’s amazing to have the opportunity to shape the organization of the company. But at the end of the day you really have to break it down to, what is the problem you’re trying to solve and how do you build up a team with that?

For example, we have three basic teams in our company, big teams. One is operations: how do you think about day-to-day operations, how do you make sure you’re plugging in with all these different vendors correctly so the data is coming through you and then you can monitor the progress and see how things are working and everything around that. Next to operations, we have another function that focuses on how do you build relationships with providers across the company to make sure they join our program. And, lastly, it’s data science and clinical: how do we then measure provider quality over time, give feedback that is relevant to them, et cetera. So if you break it down really simply: teams are really built around what are the problems the company is trying to solve. I’m oversimplifying, obviously, but ultimately, teams shape up to that.

Hiring is definitely the hardest part of our job; it’s a problem that every company faces. I think one of the hardest things is finding a team that is able to work cohesively next to each other but just as challenging is about hiring the right person at the right time. It’s thinking about when do I hire, how to hire, and is this the right hire? Is it the right time, the right person, the right skill set? How does a company make sure that the next person they bring on is going to be an overall win for the company? Oftentimes at a small company, it’s also about asking how do I make sure that if my business changes all of a sudden, my employees can be flexible enough to do something else and still add value?

That’s why at a startup you often hear there are a lot of “generalists.” Everybody is a jack-of-all-trades; they can do anything just to get an idea off the ground. Then later on as you grow, you hire people who are more skilled and specialized in a specific area. That’s something where we found, as we’ve grown, as we’ve gotten more traction in the market, we have to bring in people who are more skilled and experts in doing one thing—for example, in enterprise sales or Bayesian statistics. The ethos behind hiring changes as a company grows.

So the long answer to your question, it’s really changing on a day-to-day basis. You have to be very thoughtful about how you set up a company structure. Bigger companies like Accenture—it’s a completely different story, where there are 300,000 employees across the globe. One of the things that we very much screen for is culture fit. Everybody says that, but it’s really worked for us, and I think it’s very similar to the Orientation Week advisor selection process [in college] where you not only have to bring in somebody that you can make sure that they get the job done, but beyond just that. It’s around how do they interact with other people in the company and if they fit with in the mission-oriented focus of the company.

Being able to prevent misdiagnoses in data science is already a monumental feat in healthcare; that’s something Covera does. Do you think that there are any other areas in healthcare where a deliberate data science approach could really move things forward and have a really big impact?

Absolutely. That’s why you see so many healthcare startups in the market today. Everybody is applying data science and data analytics. I’ll just talk about Covera Health since it’s closest to home. Healthcare is huge—there are so many different facets of healthcare, e.g., pharmaceuticals insurance, providers, et cetera. Healthcare is a huge industry, and it’s been very slow and stuck in the past because a lot of the systems were built then, and there has not been much incentive to change things. In the US, the system is notoriously cumbersome and slow. There are a lot of companies that are really focused on how to reduce waste in healthcare. There’s a notion that there’s a lot of unnecessary care happening; a lot of inappropriate care happening. From a nonclinical side, there’s also a lot of administrative waste.

For Covera, the approach is to say, how do we think about healthcare quality through a very objective, data-driven way? It’s really hard to measure quality in healthcare. How do you say Doctor X is good, Doctor Y is better, Doctor Z is the best? You can get three patients and they say, “Oh, I like Doctor Z because he has very good bedside manner.” And he/she gets five stars on Yelp or on Zocdoc. While we think that is important, for us, what we want to focus on is the quality of what the doctor is trying to do.

In our world we currently only focus on advanced radiology. Think very complex images (e.g., MRIs, CTs) that patients are getting, a consultation to get a diagnosis. This is extremely important because this diagnosis basically cascades downstream to what your treatment path is.

The problem that we’re trying to solve is how do we use data to identify the probability of when images are read inaccurately. Advanced imaging is very complex; the radiologist is looking through hundreds of images back-to-back. We use the power of data science to identify quality of diagnostic imaging and help route patients to the most appropriate doctor so that they can receive the most accurate diagnosis.

On the topic of data science, why do you think this field is so up-and-coming currently? Rice University is adding a data science minor next semester, and big data is a buzzword that gets thrown around a lot. What’s your personal take on data science and its trend right now?

I think data science should be a requirement. One of the biggest things I regret not studying in college is taking more advanced statistics and computer science classes. I recently completed online courses on basic coding and applications in data science because I think it’s extremely useful in this day and age. I can’t have a tangible discussion with data scientists at my company if I don’t speak their language. I really think people make it harder than it actually is to learn the basics: I truly believe that the hardest part is to be able to think logically. Coding is like learning a new language; it takes some time, but it’s not hard if you know how to approach it. I think the ability to think in a structured manner with data is super important. That’s why I think it should be mandatory.

I think the reason why this trend is happening right now is because one, we’re turning a corner where we have so much data in the world, whether it’s in healthcare, social media, banking, transportation, et cetera. Every single Uber ride is a data collection point. Every single thing you do on your phone is a data collection point. We have these new means of collecting data at an unprecedented pace. I think there’s a crazy statistic, something like the data collected last year alone is more data than the data collected the last 30 years put together, or something like that. That’s ridiculous. The ability to collect data and store it is really changing the way we’re thinking about the world.

Number two, the data we’re collecting is being used to train algorithms to do certain things. Even though the concepts and the math behind these algorithms may not be new, we’ve basically been able to process the data at a rate that we’ve never been able to do before. That’s really shifting how the world works.

Interview excerpts have been lightly edited for clarity and readability and approved by the interviewee.

Akın Deniz Heper
Yale-NUS College
Singapore (1.3° N, 103.8° E)

featuring Anıl Korkut, Assistant Professor, Department of Bioinformatics & Computational Biology, University of Texas MD Anderson Cancer Center, Houston (29.7° N, 95.3° W)

Anıl Korkut is an assistant professor at the University of Texas MD Anderson Cancer Center. On a general level, his work concerns computational research in cancer; specifically, he runs the Network Pharmacology Lab and works to utilize network pharmacology and computational models to predict and avert resistance in cancer therapy. His research focuses on finding causal links, mostly on the proteomic level (i.e., using the protein networks of cancer cells), and developing tools using computer models to determine appropriate therapy approaches to various kinds of cancer.

My most important takeaway from my interview with Dr. Korkut was the significance of passion and excitement in science. He has dedicated his academic career path to the pursuit of science, beginning in high school with chemistry class, followed by a double major of chemistry and molecular biology and genetics in college, and then a PhD at Columbia University, where he realized he was going to work in biophysics. He told me he decided he wanted to focus on computational biology while he was dissecting larvae as a part of his first project in the US. I related to this deeply, though in the opposite way; I realized I did not want to work in computational biology when I did an internship in a genetic diagnosis center and found myself trying all I could to remain in the PCR (polymerase chain reaction) section and away from the bioinformatics lab. As Dr. Korkut pointed out, however, the future of science, and of biology specifically, is in the merging of computational and experimental methods. A part of Dr. Korkut’s work, accordingly, is in building computer models for cancer research and therapy. He believes that our growing understanding in cancer biology, coupled with patient profiling methods and computational models, could lead to cancer becoming a manageable disease in the next 5 to 10 years.

Dr. Korkut’s position at University of Texas also extends into the managerial side of education. He is critical of the way PhD programs prepare students for work environments because of the focus on academia and tenure positions despite the vast majority of PhD students moving onto the private sector—sometimes to positions unrelated to research altogether. He talked of a need to reform the way we approach science education, which spoke to my own experience. As much as I enjoy studying biology, I feel constrained to a handful of options in going forward with a science education, which does not reflect what options are actually available. The career pathways that lead from an education in science are numerous and growing each day, from pop-science journalism to research in the private sector, and yet the majority of science programs promote a direct path to academia.

Our discussion about the academic path for scientists touched on one of the largest problems in academic positions in science, which is the way success is measured. Currently, the profitability of a study, how many publications a researcher produces, and the impact factor of the published journal are used as metrics in determining the success of a scientist. The true value of a discovery, which may lead to a slew of other studies and discoveries, can be glossed over. One of the largest disadvantages of a career in science, Dr. Korkut said, is contending with this issue. It is also one of the downsides that has been turning me away from a future career in science. I do, however, find learning deeply enjoyable. I resonated with Dr. Korkut to a great extent when we talked about the most exciting part of his work and he told me about the feeling of knowing something that nobody else, ever, has learned before. That feeling, I believe, is what drives most people into an education in science and is one that I hope I will be able to achieve some day.

My interview with Dr. Korkut was very educational, and I learned a lot about the computational part of cancer research and the pleasures, as well as the pains, of working as a scientist. I found some of my fears about a future career in the field of science, and biology in particular, affirmed, as he emphasized the necessity of passion and commitment in his field. The dedication demanded by such a career is substantial; the rewards of knowledge, of discovery, and of invention, however, are unparalleled. Personally, this dynamic of great struggle and great reward has been the source of my doubts in furthering an education in science. The international nature of science, the capacity to collaborate with people from every corner of the world, and the opportunity to work for the single purpose of discovery has been pulling me back at each turn. For now, as I strive to find my passion, I will remember Dr. Korkut’s statement, “Science is not an afterthought.” In my career, I hope to make sure it never is.

Highlights from the interview

What made you say I’m going to get into science and biology?

That was really an early decision that goes all the way before starting primary school. It was almost obvious that I would choose something in science. We used to have a big encyclopedia in our old house that was called World of Science. My father used to read chapters from that book, and I was always fascinated.

Then I went to a special high school for people interested in science, and there I started doing a research project. My main interest in those years was in chemistry and a little bit biology. Later, I made a decision to choose physics. But then there was a new emerging field, molecular biology and genetics, that caught my interest in college. After having discussions with university professors, I decided to get into molecular biology and biophysics. But from the very early periods in my life, my main goal was doing science and contributing to science. It was a great excitement for me. I deliberately avoided going to engineering school; it was really about basic science for me. I was exposed to all that education in my home country, Turkey. Then I was lucky that there was a professor who came from Harvard to my department, and she suggested that I spend a summer, maybe a year, at Harvard doing research. So I came to the United States, and I started working for a project in neurodevelopment. My first project was dissecting fruit fly larvae to look at neuro-developmental patterns, and I quickly realized it wasn’t the right fit for me. I realized that I was interested more in computational science, and I was very lucky that I moved to another laboratory at Harvard doing more quantitative biology, structural biology, which I decided was a good fit for me. Later, I came to Columbia University for my PhD. At the time, it was obvious I was going to do something in biophysics, and from that point on all my career focused more on quantitative sciences, and then moving to more cancer biology, and after the postdoctoral period at Sloan Kettering, I came to MD Anderson.

Did you have a mentor or anybody you looked up to that led a path for you?

Yes, indeed. I can say I was extremely lucky in having very good mentors at all steps in my career. The first advice I would give to young people is you’ve got to find good mentors. It’s very important. And you need one at the early phase in your career, even if you become president of a university one day, or a Nobel Laureate, or CEO of a major company. You will need a mentor. I had a couple people in different departments or institutions, and my PhD advisor was a great role model for me. I was lucky in that sense. There’s no one recipe in choosing a good mentor. You first need to understand what you want to do in life, and then you have to find someone who achieved that, and in a way that would be compatible with your nature, and then you need to get mentorship. There’s not a single optimal mentorship recipe though.

What do you do in your current position?

There’s a technical and scientific aspect, there’s a managerial aspect, and there’s also a mental part to my position.

On the technical side, I am running a research laboratory with about six, seven people. Half of the laboratory is developing algorithms to analyze genomics data from cancer patients to predict how a given patient will respond to therapy and whether the person will be resistant to a conventional therapy and what can we do to nominate new and effective combination therapies. We also have an experimental lab where we do a lot of molecular profiling of cancer patients using imaging and other methods to validate our predictions from computational models. Basically, getting the right cancer models in the lab environment and testing the drug combinations that we predict that should work based on the computations. This is a highly important problem because we have new generation drugs that work reasonably well in cancer patients, but response is not durable, so our goal is to really find better drug combinations that will give more durable responses. Ultimately, we want to contribute to turning cancer into a manageable and curable disease, where you simply identify drug targets in a given patient at a given time—it’s a dynamic process—and come up with the right therapy at the right time so that the patient will live a healthy life. So that’s the technical part.

On the managerial side, I am basically running a laboratory, and that’s the most important part. We have a lot of young scientists and some more senior people as well. I’m blending these people to run projects in different research areas, which could be biostatistics or cell biology or physics. And my real goal is to (a) channel them to develop and solve scientific questions, (b) showing them how they can likely be more successful and also make sure that they have the right—enough—funding. A lot of time that I spend is writing research grants and research papers so that we get federal and private funding so that we can pay people’s salaries and also cover their research expenses; that means acquiring an enzyme or sending them to a conference. So I’m responsible for that too. That takes a lot of time. I also handle the collaborations. We have a network of collaborating research scientists. We act together, we write grants together. So I am coordinating all of these collaborations. That means people who are doing similar research but also complementary of course. I network with these people, we come together and we brainstorm. Some of them are at MD Anderson, and some of them are at other institutions, so I handle all of those partnerships.

And then I have a little bit of administrative roles that I try to minimize, of course. Sometimes it’s fun, like participating in recruitment of new faculty members. You have to read their papers and CVs and interview them. Or we recruit other postdoctoral researchers, so I participate in those committees, and I also participate in the university senate. Once a month we discuss the problems and new opportunities in the institution. And it can be anything, like something about data security to research funding or rules.

On the research side of things, I realize that you mostly focus on the proteome level in your research. What’s made you gain that focus?

That’s a very interesting question. It started that way, but it’s not exactly true anymore. In some sense, yes, I still think the proteomic level is the most important because it’s really reflecting how a given cancer cell will respond to some of the state-of-the-art therapies, and proteomic levels are critical for predicting whether a cell will be resistant or not. But then there is another level that is the genomic (DNA and mRNA) level that also matters. The genomic alteration that means mutations and copy number changes, mRNA expression and epigenetic changes, in some sense, define the infrastructure in the cell—determining the potential, the capabilities of the cell, but it’s the proteomic level that defines the refined plasticity. And the response to therapy. So, in fact, we need to understand both parts. If you can get away by simply understanding how the proteomic landscape is, you can still come up with therapies. But you usually need some genomic information to figure out how a cell may behave in response to therapy. I actually focus on both.

In my research, one part is team science. And that’s like the team projects such as TCGA, which recently ended. In that case, actually dozens of laboratories come together, and we do a lot of genomic and proteomic analyses of large patient cohorts using all sorts of data on protein activity, DNA mutations and mRNA expression. That involves a lot of computational analysis. In the lab, we use mostly proteomics, and right now we also look at image-based proteomics which gives us a lot of information about tumor microenvironment; that it is an emerging field. We understand now that the tumor microenvironment is very critical in determining and conferring resistance to therapy. Many new drug targets are indeed in the tumor microenvironment, but proteomic measurements for tumor microenvironment are challenging yet really fun to do, and also it’s very promising in terms of finding new drug targets. But it’s a mixture of things. There’s no one recipe. Sometimes you look at genomic levels and sometimes proteomes. 

And what can you tell me about perturbation biology?

Let’s first start from a system where there is no perturbation. We can make measurements on thousands of molecular identities, be it gene expression, protein levels, and all other things. And you will see a fairly rich landscape full of molecular associations that things melt together, go up or down. Most of these are associations and quantified as correlations, yet you know very little about the causal links in the cell or how cells will response when you intervene. So there are three levels, the associations, the causal links, and response to interventions on the system.

Perturbation biology gives us a tool to extract causal relations from such associations because you hit the system, you change one parameter, and then you see how other parameters changed together with that. And then you do a perturbation in the opposite direction, and you observe how your system changes and whether parameters change. And doing this in a repetitive, in a rich perturbation setting, the causal links emerge. And if you can indeed combine, integrate all those causal relations and put them into a systems framework such as a rich network model, then a new pattern emerges, and you can start predicting how the system will respond to previously untested, not applied perturbation. That gives us a chance to predict how we can design, let’s say a therapy, a new therapy that was not tested before. So, perturbation biology is helping us to form associations, molecular associations, and start to build these causal links and using those causal links at the higher level, in some sense, to predict how the system will respond. That is the perturbation biology. Actually, this approach is not unique to biology. In any science, to understand how a system will behave, you need to perturb the system and watch how it’s evolving.

And can I also ask you about network pharmacology? 

The network pharmacology—some of these terms are newly emerging. There are few leading groups like Peter Sorger group at Harvard Medical School, and Chris Sander’s group, again at Harvard Medical. Gordon Mills, who was at MD Anderson, now at OHSU, and Joe Gray at OHSU and couple other groups, that we coined these terms together. There’s, for example, a Laboratory of Systems Pharmacology at Harvard, and there is another group called Center for Spatial Systems Biomedicine at OHSU and that does a lot of perturbation biology. These are emerging themes in modern biology. Network pharmacology or systems pharmacology or quantitative pharmacology in this context can be seen as a subset of perturbation biology because it really focuses on how pharmacological agents can perturb the molecular networks within a cell. And those networks are indeed affected and rewired by extracellular or internal perturbations. The endpoints of these effects are the phenotypes such as cell death, survival or proliferation. All these entities are connected to each other. You apply the perturbation biology idea as you model the cell as a network and use pharmacological perturbations to interrogate the network, and that is what network pharmacology is. And you can actually do this using different small compounds such as clinical anti-cancer agents or new technologies such as CRISPR.

Are there any large misconceptions about what you do, both on the research level and on the managerial level?

This question could be answered in many different ways. I’m a relatively young scientist at this level. I think one thing that is misunderstood about science these days is success of a young scientist, or even a senior scientist, is measured by how much funding you bring or how many papers you write. And those are used as metrics by funding agencies or university administrations, et cetera. I think the success in science is very hard to measure…but the real focus should be more on the discovery, finding new things. Not, you know, how much money you bring to the center, or how many papers you wrote this year regardless of the content, or what is the impact factor of the journals, where you publish. These are artificial things that we create in the contemporary context. But a hundred years from now, none of this will matter. What will matter is the groundbreaking discoveries you make, and we need to think a little bit more on that. We sometimes lose the context, in that sense, for what is important. That’s one thing, I think, within universities we start to lose that, and sometimes there’s another misunderstanding from the society. People think that we are able to bring quick and definitive solutions, say, we should be able to cure cancer in a couple years, but in fact this is a very complicated problem, and there…bringing a cure just by doing a small research projects is not easy. We need a lot of team science to achieve this goal. And sometimes people expect simple recipes, which do not exist. Having said that, we are entering, I believe, a very exciting period. With all the understanding of the genomic landscape and the new tools we have, the ability to profile patients with multiple biopsies and profiling how tumors will respond to therapy, we will probably be able to turn cancer into a manageable and curable disease. I am very optimistic but that takes time. 

That is a very hopeful thought, to be honest. What do you think is one of the biggest disadvantages in your field? Do you have moments where you feel like “I should have been an engineer” or maybe that was a better idea? Or nothing like that?

Let me put it this way. With all these advances in medical sciences, antibiotics, heart medications, et cetera, still a healthy person lives around eighty to ninety years at most. Sometimes it may take 20 to 25 years, a substantial fraction of one’s life, to finish a project and make a real high impact discovery. In my field, sometimes it really takes a lot of time to see how your hard work pays back. So during the process when you do something, you fail, do again, fail again, do, fail, do, fail. There are points when you say, well, I could finish another school, find a good job, can become a manager when I’m 30, and just live and wait for my retirement. In science, there is a long period when you are kind of in the dark. So in those times you sometimes say, well I could have a different life. But now when I look back, I can see the future a little bit better and become more optimistic. I see a lot of young people suffering in the same way. Indeed, a lot of my friends had to quit at some point because I think, like in many fields in the world now, the competition is harsh and return of hard work is not clear.

What reconciles all of this in your job? What makes you say, I’m very glad I’m still doing this?

There are times—let’s say once in every couple years, and it’s getting more frequent as things get more established—there are points that you learn and know a certain thing that others don’t. Learning is always fun. For example, you read an article or open a Wikipedia page, you learn something, you find it interesting, and you enjoy it if you like scholarly things. Now imagine that you learn something that you know no one else knows. And you’re the very first person, and then you know that you’re going to start announcing it, and people will find it very interesting and will just enjoy that, and one day it could even turn into something useful for other people. I think that’s the most fun part. And second, you see other young people coming and trying to learn from you, and you train them, and they also pass the stages you had passed 10 years ago or so. Seeing that is also really fun. 

So going back to changes in your field, what do you think the future holds for cancer research at large and specifically in your case?

There are similarities to other fields. There will be more automation and more predictive algorithms, and things will change, just like driverless cars versus today’s cars. There will be similar changes in science too, more automation, et cetera. But I think the most important thing is the boundary between experimental and computational work will disappear, and we will completely be integrated with robotic devices, but the algorithm development will be extremely important. We will have a brain machine interface to make new discoveries, and things will get very fast in that sense. The data sets will become very large but, as a colleague of mine just said, the big data in biology—I’m sure you know the term “big data”—is not really big today. Compared to click data or other marketing data, et cetera, our data sets are limited. This is going to change quite rapidly. And a lot of other big changes will follow. Another thing is we will, of course, have better profiling methods from patients. That actually is one of the factors that will change the volume of data, but that also means we will be able to predict the immediate molecular changes in a given patient after therapy, and that will enable us to come up with better therapies especially drug combinations. And I think we are going to change people’s lives. There’s definitely a shift from more theoretical, or basic science in biology to more applied and predictive projects, so I’m expecting, also, a commercial explosion in that sense. There will be a lot of biotech companies and, in some sense, more flow of private funding into biology. There are good parts of this, and also there are some challenges, of course, because we have to keep quality high. You cannot create bubbles in life sciences because this is not a bunch of software; biotechnology affects people’s health and lives. So that’s going to be important; we just cannot sell snake oil. There is that danger, of course, if there are tens of thousands of new companies emerging. We have to be careful.

What advice do you have for interested students? Just in research in general and specifically in computational and cancer biology?

I will give the same advice to both. If you really like it, and if you have an excitement, do it. There’s no question. Science is not an afterthought. It’s not like a second choice that you just had wanted to do that, that didn’t happen, let’s be a scientist. Or I don’t know what to do, let’s do a PhD. Don’t do that. You won’t be happy. There are cases that people really become more interested in time, but if you have a passion for science and learning new things, don’t give up just because there are challenges. But also if it seems like the only option for you, or you just don’t know what to do, don’t do it. So this is for people who really like it. Also for people who want to do cancer biology—I think this is the right time, and there are so many good programs, I believe there will be a lot of opportunities in the next 5 to 10 years for people in this field. Landscape is quite promising, I can tell you that.

Is there anything else you wanted to mention that came up during the interview and I didn’t ask?

One thing I want to say is there are really three things in life that matter, and that’s really special, and I think it’s science, arts, and sports. Science is one of the few things that is truly international, or global. I mean, it’s not even international. It doesn’t really care about national versus international or intranational, so in that sense these are truly universal values. If you want to do something special in your life, I think science, art, and sports are really great things. Even if you don’t become an expert or a professional in one of these fields, I would recommend every young person should enjoy at least one of these things. That makes your life much better.

Interview excerpts have been lightly edited for clarity and readability and approved by the interviewee.

David Yang
Rice University
Houston (29.7° N, 95.3° W)

featuring Pierre Elias, Cardiology Fellow, Columbia University Medical Center, New York (40.7° N, 74.0° W)

Pierre Elias is a cardiology fellow at Columbia University. He completed his residency training at New York-Presbyterian Hospital/Columbia University, his MD training at Duke University School of Medicine, and his BA at Rice University. Along with practicing cardiology and seeing patients, Pierre conducts research in health informatics in order to make better use of existing healthcare data, and he works with the New York-Presbyterian information center and venture capital groups to improve technologies in the hospital.

When I was preparing for this interview, I had expected to talk with Pierre mostly about informatics and cardiology, two areas that were new to me. The interview, however, introduced me to a new mindset of approaching college and life afterwards. Pierre first discussed how students in college tend to strive for clearly set paths that have prestige associated with them (e.g., medical school, law school, or consulting) that all other students also seem to be striving for. I was impressed to see that another way to approach college and post-graduate plans is to figure out what problems I am interested in working on, and everything will follow afterwards. The problems I work on can change over time, but passionless work is to be avoided at all times. After the interview, I thought about the involvements I am engaged in and if they were issues that I was passionate about. I think that is a good self-assessment to take every now and then to make sure I am not just doing what others are doing in order to “fit in.”

Another part of the interview that impressed me was the role of data in medicine. We live in a world that is becoming increasingly reliant on data and information that is generated faster than we can process and digest. In any field I go into, there is the potential for enhancement with learning programming and data analysis skills.

After the interview, I was inspired to follow my interests wherever they may take me, and in order to do this, it is important to be exposed to issues and perspectives that can spark interest and passion. This can occur by reading new things on the internet, attending conferences, or even just emailing someone who works in a field that is interesting to you. I hope to take the lessons I’ve learned from this interview with me after I graduate college, when routes may become less structured as I decide on how to spend my life on something meaningful to me.

Highlights from the interview

I am a junior. I am a chemistry major, and I know I am interested in healthcare, but I am very open to whatever capacity that I can be working.  Healthcare is a huge field so I want to explore the different aspects of it.

So, you like science. You want to help people, and you are trying to figure out if you should go to med school, because everyone is going to either end up in grad school or law school, med school or being a consultant after college. And going to med school is kind of a logical and prestigious thing to do, so you are trying to figure out that is the right thing to do. Does that sound right?

I am eighty percent sure I am going to med school but basically you are correct, to know for sure that this is the path for me.

My goal is to try to knock that certainty down like fifty percent by the end of the interview.

Oh cool!  That’s good.

I think Rice takes in these really smart people, and it shows them these paths, and does a great job training them, but at the end of the day I always felt like there is this overemphasis on—I don’t even know if it is coming from Rice or students or what—these kind of very clearly set paths, like, I am going to go to grad school or law school. And there is a lot of comfort in that. The problem is, that is not how the real world works.  There are ten professions you see on TV, but there are a thousand different jobs you are not even exposed to. The truth of the matter is, if you are going to graduate from Rice, and do well at Rice, you are never going to go hungry again. So it is easy to feel like you have to go down these paths.  But oftentimes, the most interesting jobs are the ones that you never hear about.  

Medicine is this super oversaturated field. Even if you want to make healthcare better in this country, there is a question of—is going to medical school the best way or the only way to do it. In my undergrad years I was definitely underexposed to some of the other opportunities that are out there. I want you to feel like, okay, I settled on going to medical school, but at least I knew three other things that I could do that weren’t going to medical school that would be a great opportunity.

Do you at all regret going to medical school, or are you perfectly happy with your decision?

I don’t regret it, but I think the reason I don’t regret it is multifactorial. One, I chose where I went to medical school because of the opportunity afforded. I went to Duke for medical school, which has a really unique program. The traditional medical school is two years at the books, two years in the hospital. What Duke did is they took these first two years and they made it one year. It is very jam-packed, an accelerated program. It is eleven months of preclinical, which is just brutal. Then your second year is in the hospital, and then your third year you are off. So, basically, you learn the books, you go to the hospital for a year, and then you are off—free to do whatever you want for a year. That was an incredible opportunity. Trying to decide where to go when that was offered to me, I was like, I want that because I want to explore a bunch of other stuff. I was choosing between Duke and UCSF for med school and ended up splitting the difference because I went to Duke for med school and out to UCSF to do my research.

Nice!

[To decide what to do during my free year] I would talk to people about what I should do, and one person in a phone call told me to talk to this other person, and that person said I should talk to a third person, and then the third person was Bob Wachter. I never heard of him before. It turns out he is a really big deal in his field. I had a couple of conversation with him. He said, “Why don’t you come out and work with me?” After two phone calls I moved my life to San Francisco. I had this incredible time at UCSF, and I also had this incredible opportunity to work at a tech company, first as an intern, then as a data scientist. I liked medical school. After I worked full time at a Silicon Valley startup, I decided, you know what, I do want to still practice medicine and came back to it.  

Medicine is far from perfect. The process is long and frustrating. There are days when I am extremely annoyed by it, but I think the reason I don’t regret going to medical school is I went for something that gave me opportunity to explore other things. I am going to ultimately find a way to find a hybrid of all the different things I did and want to do. I think I would regret it if I never had the opportunity to leave, and try something else, and figure things out for myself. That flexibility is really, really important. 

Awesome. So would your advice to undergrads be to basically explore as much as possible?

My advice, when I talk with people who are trying to figure out the next step, particularly if they feel like it is a big next step, is expose yourself to problems. Real problems. And really see what it is like to be in the day-of-life of different people. I think the more you expose yourself to problems, once you find a problem that matters to you, the rest becomes so much easier. You worry less about posturing and think, oh, should I go to medical school or grad school because it is prestigious, or looks good, or should I go to this place because that looks good…Yada yada yada…You just say, this is the problem that matters to me and in order to fix that problem, I need to have these skills. How do I get to a place that is going to give me those skills? And you realize that all of these things, whether it is medical school or taking a job right after college, oftentimes what you are given is a very broad, boilerplate version of something, which is less than impressive. You know, it is so broad that it is hard to feel like you are becoming an expert in anything. It is going to be really hard to decide what are you really dedicating yourself to. You have to say sooner or later that this is the stuff I am going to learn and this is what I am good at. That is incredibly difficult when you are given a ton of options, but if you know what it is that you are trying to achieve—particularly if you know the problem that you want to be a part of fixing, your understanding what your value is, and you have a framework around how you make your decision—that, in the long run, is a lot better. When you look back, you will say, “I made those choices. It wasn’t because I thought this was the best thing, or I thought it would look like the best thing. I made that choice because I trying to fix something in the world, and this appeared to be the best way I could move forward.” I think prospectively; it is helpful because it helps provide some architecture to decide how you are going to make decisions.  

How did you come to know the problem you wanted to fix, the problem you were interested in? 

Problems I have worked on have definitely changed over time, but for me it really came from that exposure of my time in the hospital. I thought healthcare was interesting. Okay, med school seems like an okay place to go. And then I didn’t really like my preclinical time very much, but I don’t think really anyone does. It is just a ton of information being shoved into your head as fast as possible. Then I got to the hospital, and I liked seeing patients, but for me it was looking at the way—how it works.

Basically, every day in a hospital, life and death decisions are being made. The people who are working the hardest, like residents in the hospital, have to overcome insurmountable logistical problems. You will watch someone who is working eighty hours a week trying to take the best care of super sick people, then spend thirty minutes on the phone trying to make sense of a project after that. I looked at this mismatch and thought, this is insane. Or you pull out your phone, and you get to use these incredible technological apps on your phone that work seamlessly and then you have this piece of crap ten- to fifteen-year-old software at the hospital to make life and death decisions. So, for me, it was really about—why is the technology that we use in healthcare so bad compared to the technology we use in the other parts of our life, even though the stakes are so high. That is what, for the last eight to ten years, I have been totally fixated on. I can’t accept the way things are around that. You know, I have been working on these problems, and thought about these problems, and work with the best people in the country around this stuff, and still, every day, I am mad about the quality of technology that we have in healthcare. It is something that pisses me off every single day. If it is going to be something you dedicate your life to, or at least dedicate your career to, you should be really passionate about it. This just became clear to me that this was something that I was really passionate about. 

When did you first envision yourself as a physician, clinical researcher, or a cardiologist?

That is a good question. I was interested in medicine in high school. When I went to college, I was taking pre-med classes, but I was also a sociology major. It wasn’t until junior year I thought I may give medical school a try. For me, the way I made that decision at the time— if I try to look at everything bad about being a doctor, and if I am okay with everything bad about being a doctor. Every job has perks, but the downside is we undervalue…I think if I convince myself that I was okay with all the things that were negative about being a doctor, then I knew I could be especially happy in the area. I really had no idea that I wanted to be a cardiologist until I started practicing medicine.

Did you have someone who acted as a mentor to cultivate interests?

I had tons of mentors. One of the big things I learned is that there is no such thing as having a single mentor who answers all the questions. Frankly, mentors can give you really good advice in one area and really bad advice in another. Over time, you have to learn.

Most people don’t know for certain what job they want to have, and also a lot of times the job that is going to be perfect for them does not exist yet. It is going to be a hybrid of different things. It is really hard to meet someone and say, “I want your exact job.” And even if you do meet someone like that, you don’t really know how they got there. It changes. I think that is the advice I would give about mentorship, have multiple mentors and don’t necessarily assume everything you hear is going to be right. Get different opinions and think for yourself to figure out what is it that feels right, what is it that doesn’t feel right. And then make your own decision.  

There is no point in asking people where do they think they are going to be in five years. There is so much that is going to happen between now and then that that answer is going to change. We [all need to understand] that people’s careers are representations of them doing good work around companies and being presented to you as opportunity—and moving left and right through this opportunity, and that is completely okay for yourself. You have to look at your local environment and work on something that matters to you and realize you are going to float through the river of opportunities that are made for you. That is how most people’s careers are and that makes it less stressful.

You mentioned that in your undergrad that you majored in sociology. Why did you choose to do that, since it is a nontraditional pre-med major? Do you use that major in your day-to-day work life at all?

I think I ended up being sociology major because during orientation week I met a really nice girl. She told me that she was taking a sociology class. She was pre-med, and she seemed to have more sense to pick out classes than I did, so I ended up taking that class with her. I ended up gravitating toward sociology partially because I like the subject matter and partially because it offered me flexibility to take all sorts of classes, which I enjoyed. I could be pre-med, I could be sociology, and I could still explore other interests. That was one of the other benefits of being a sociology major.

I would say sociology changed the way I look at the world. To recognize that when you interact with someone else, a lot of how that interaction is going to go was already dictated before you opened your mouth. There are all of these social contracts that we live with that determines most of how things are going to go between you and I. And understanding how much is dictated by the way these are constructed is, I think, very profound, and it could make you super empathic. You may take care of a patient who is extremely difficult to take care of, and he is yelling at you and can be making all these mistakes with their personal health, and extremely health destructive. You can be very frustrated, and it can be very easy to entirely blame that person. Or you can recognize this person grew up with a life very different than your own. The opportunities were very different. You can start to understand maybe part of the blame belongs to him and maybe part of the blame belongs to society. Starting to understand that this person had a much higher chance of ending up here just because of the life they were given when growing up. And the way the society is structured. You get to understand the gravity, trying to make society better, as well as to understand that not all things are determined by, first of all, effort.

Can you give a brief overview of the kinds of projects you are working on? Why do you think your field is important for both physicians and the public to know about and care about?

There are three major things I work in. I am a cardiology fellow, which means I practice cardiology and see patients at Columbia, and basically do all things cardiology-related, whether it is noninvasive cardiology, taking patients to the CATH lab, or cardiac infection care. So that is one part of what I do. Seeing patients, learning about cardiology, practicing cardiology, and getting to delve into all the things we do taking care of people who have cardiology problems. So that is one part. Then there is another part, which is research, particularly around health informatics. I am very interested in how do we make better use of healthcare data that exists. We are at a very interesting inflection point where we greatly increased the amount of data, simple data, that is available anytime, but we still don’t know how to use it. Data is information, information is knowledge, and knowledge is wisdom. We are at a point where we greatly increased the amount of data that is being generated, but we are still very far away from turning that into information or knowledge so they can be used in the field. So, how do you unlock that puzzle? That is a big part of what I work on. I am very interested in the ways we can make technology automatically work for us. Machine learning—specifically machine sensors, so things like machine vision—to better automatically interpret images and get better interpretation of clinical imagery like echocardiograms. So that is one part that I work on, pure research side. 

And then there is another part which is the more practical side, which is we know we need to do a better job providing high quality care at a low cost in this country. We have a very inefficient, very expensive healthcare system. There are a lot of companies that are working on that. So how do you work with a technology startup company that has great technological expertise help solve these problems. I work with the NewYork-Presbyterian information center and venture capital group—and to help find companies that we think are doing important technological work in healthcare and bring their technology to the hospital and employ that technology in the hospital, trying to improve the clinical care that we provide. Working on some of those technological solutions and figuring out how do you actually take a grand high idea and make it work where the rubber meets the road. You know, how do you deal with potential problems taking new technology and making it work in a highly complex, highly variable, and highly risky situation.  

For research, you talked about using machine learning, dealing with data, and working with tech startups—how did you receive the training to get you to this point? Was that what you did in your third year off from medical school?

That is a good question.  I would say a huge portion of what I need to know in order to do that work would not be found anywhere in a traditional medical curriculum. And that is not necessarily the fault of medical schools. Medical school is supposed to train and prepare you to be a doctor. And I say that to get to that point it is very optimal. Very likely you will end up doing more than one thing in life, from a professional standpoint, and that training may not necessarily be from professional schools, grad school or undergrad. A lot of learning comes from on the job or finding it yourself. The way to figure out what it is you need to learn is to find a problem that matters to you, to understand how you are going to approach this problem. 

What started for me was understanding the problems I wanted to work on and going to people smarter than me and saying, “This is important to me,” to see if they say “Yes, that is an important idea, and that is a good thing to work on.” And then say, “What is it that you don’t know and you need to know?” So it would be things like recognizing that you need access to the data, you need to understand the way the data is structured. For example, Epic has one of the largest electronic health records in the world. So they sent me to the headquarters for Epic, and I spent two weeks there learning about the way the data is structured and getting certification of data. And then I would start with little projects, and work my way up, and start to learn about how do I actually do data science, how do I learn how to do extraction of front load, how do I actually understand the data. And when you are actually given a problem, and you need to answer that problem, you learn at a much more exciting and engaging pace because you have a very clear reason to learn something. And you need to learn it right, so it gives you an excuse to delve more. A lot of it was: I have a problem, I need to answer that problem with the resources that would help, and then I went about trying to answer that problem.

I also expose myself to different experiences. I kept looking out for different opportunities, and kept putting my hand up, and kept trying to do good work, and kept asking for more.  I can give you an experience where I did something nontraditional, but it worked out incredibly well. There are three major TED conferences each year. They are TED, TED Global, and TED Med. I loved TED talks when I was an undergrad. I learned more about TED Med, and I called to see if they had scholarships. They were offering colleges to just basically—to those who go to TED Meds. I filled out the college application saying I would like to come to it. This is what I am doing. This is why I am so excited about this. And I would literally like babysit your kids, please I would love to have the opportunity to come to this conference. I was lucky enough to get the scholarship, and I got to go to TED Med at Johns Hopkins. It was a great experience. I talked to every single person I could. And I made a first connection. After that, they said, “We need help. We get a thousand applications for speakers at TED Med every year. We need help vetting for this, would you be able to do that?” I said I would love to do that. I got to select the next speaker for TED Med and they liked how I did that so I got to vet speakers. They sent me to TED Meds for two more years in a row. I went to TED Med three years in a row. I was studying speakers, I met bunch of the speakers, and I was nominating speakers. The startup that I ended up working at was through meeting someone at TED Med.  

What advice would you give students who are still looking for problems that they are still interested in? I think you mentioned that you have to push yourself out there to expose yourself to various issues.  Is there any practical advice that you would give to a college student that is still trying to find what they are interested in?

First is, you don’t have to find the perfect answer. You just have to find the answer that is going to work for now. Everything is about durations. Start with what you know you are interested in, or think you are interested in, and just go and start talking to people. Professors’ office hours, look at the TED talks you like on the internet, read about it often, and just go where the problems are and try to understand what is an important problem that needs to be worked on. And then be convinced that this is something that is worth your time, and it is exciting. So start with that. Conferences, professors, just interesting things via internet. There are so many ways to start to find problems. And really think about these problems and try to understand them. Engage with people about them, say things like okay, I do think this is important, I do think this is unsolved, and I want to work on it.

There are times when you end up working on things people are not going to support, and it ends up being very, very important. Probably the poster child for that is E.O. Wilson. He basically discovered pheromones. But he always thought it was an important problem. And so you should work on a problem that you think is worthy of your life, because you are going to be dead in something like 960 months. It is not a lot of time that you get. You only get one go at it. You should only be working on stuff that is worthy of your life. And that means working on a problem that is going to take a generation to find it. So you should absolutely look at all these problems and really— this is big, this is something that is going to define the way we live our lives. This is something that is just so incredibly important. It has to be important to you. It is okay if it is not important to others, but it has to be important to you. So talk with a bunch of people, get yourself out there, find a generation-defining problem, and then start asking yourself, “In order to answer this problem, what do I need to do?” And then you are off to the races. I have always preferred that to the equivalent about…should I be a consultant, does that pay well, should I live where I want to live, does it look interesting, should I go to McKinsey or Bain, McKinsey could do this for me…all of that stuff is how you end up in a career with a lot of accolades and a lot of…it is really figuring out how do you solve a problem that matters to you. 

(Interview excerpts have been lightly edited for clarity and readability and approved by the interviewee.)