Ben Lacar

Ben Lacar, PhD, is a data science fellow creating health information technology towards the goal of integrating social care into the delivery of health care. Since completing his Neuroscience PhD and postdoctoral research, he has transitioned to industry and worked in various roles such as biotech applications scientist and bioinformatics scientist. Ben is currently a Data Science Health Innovation Fellow at UCSF Bakar Computational Health Sciences Institute, UC Berkeley Institute for Data Science, and Janssen Research & Development where he is exploring innovative data-driven approaches to improve human health and healthcare data sciences. 

Can you describe your academic and professional background? What path led you to pursue this field?

My interest in learning and brain plasticity led me to earn my Ph.D. in neuroscience at Yale, followed by postdoctoral research at the Salk Institute. My postdoc project involved single-cell sequencing before it had gained mass adoption. Experience with that technique opened some doors to work in industry. I had a hybrid role that allowed me to do some bioinformatics and data science, customer engagement, internal training, and product development. It was a great role coming out of academia because it exposed me to different facets of industry. I got really into the data science aspect of it and was eager to grow in that area. That led to a couple of data science fellowships including one where I am now.

How did you find this particular position, and what was the hiring process like? Is there a typical structure for this in your field?

The great thing about data science is its wide applicability. I looked for roles where I could enhance my data science skills while doing work with social impact. I learned about my current position (an academia-industry health data science fellowship) initially from a social media post. A recruiter also reached out and provided some encouragement to apply. As far as a structure for the hiring process, there were some similarities to other roles I had applied to in the data science field, but some differences too. A typical data science hiring process consists of an HR phone screen, a conversation with the hiring manager, a technical assessment (coding, statistics, etc.), followed by an on-site which has more behavioral and technical questions. (Of course, many “on-sites” are still conducted remotely.)

Can you tell us about your current responsibilities? What is a typical day or week like in your role?

Earlier in the project, I was devoting more of my day to reading about the domains of the project (electronic health records, social determinants of health) or learning about techniques I would need to carry out my data science project. I still learn those areas, but currently, my  responsibilities are focused more on project execution. I’m coding almost daily. I structure my week so that I don’t have more than two meetings or talks to attend in a day. This ensures that I have long blocks of time to carry out my work. It’s nice to be pretty independent but I also value the feedback I get from my colleagues.

What do you enjoy about your current job and work environment?

I feel lucky to have intellectual freedom and have great co-workers. I work from home most of the time due to the pandemic, but I’ve been able to get work done remotely even in the past. This flexibility has been nice and convenient. I’ve also gone to work on campus to mix up my routine and have in-person conversations.

What are some of the challenging aspects of your job? Is there anything you wish you had known about your job or industry before joining?

I’m not sure if I’d call this a challenge necessarily, but I feel like there’s constant learning in data science. Usually, this is a good thing. Learning helps me stay engaged. As I mentioned above, you might be learning a domain outside of the field you’re trained in (many industries are using data science now) or learning new skills (data is diverse and different questions require using different methods). The flip side of this is to ensure that you’re not taking a role or project that’s too challenging. You want to bag some wins and contribute and not constantly walk uphill.

Do you have any professional plans for the future? What are some future career paths that could open up for someone in your position, 5-10 years down the road?

I think a natural progression is to take on more senior responsibilities, including more mentoring and strategic decision-making. This could mean taking a little bit of a step back from the hands-on coding but still using my experience to help guide more junior data scientists.

What’s changing in your industry? Are there any future trends we should be aware of?

This is likely influenced by working in the Bay Area, but I feel like things are changing all the time! We’re seeing increasing adoption of data science in health care, but there’s still a lot of room to apply in different use cases. If you think about your interactions with your own doctors, I’d bet that they’re aware of, but not using, some component of machine learning and analytics in their specialty. Naturally, the data science methods that are relevant pertain to data that’s digitized: deep learning methods for different medical images, natural language processing for doctor’s notes, and causal inference since most (probably all) EHR data is observed versus generated experimentally etc.

What activities, internships, or organizations would you recommend someone get involved with to help them break into this field?

Don’t be afraid to aim for a job or internship at a company that interests you. In addition, data science programs and fellowships are great for growing your skills and your network. You can also find communities online where people are working their way through a course or book. But PhDs are also highly valued as data scientists. Therefore, take advantage of your current position and lean into the statistical methods that your dissertation affords. If methods you want to learn are not relevant to your project, find university meetups or audit classes that help you learn those methods. With most jobs in industry, this kind of dedicated skill-building time is harder to do.

Is it common for people in your field to have a scientific/academic background (i.e. have PhDs)? Can you think of any advantages or disadvantages someone with a PhD might experience while pursuing or working in your field?

Yes, as I mentioned above. Sometimes it is even a minimum requirement on a job description.

Your dissertation will provide a showcase for when you’re looking for jobs or organizations to join. Therefore, don’t just think about what’s enough to publish, but what could give your story some nuance. It might not change your findings much and it could end up in the supplemental section, but these details allow you to know the methods deeply, including their strengths and limitations. I can’t really think of a disadvantage of having a Ph.D. 

Do you have any final words of advice for those navigating these career questions? Is there anything you would have done differently given what you know now? 

It’s hard to say what I would’ve done differently because sometimes you have to focus on the job at hand and motivations can vary at different parts of your career development. Like I said above, you’re always learning in data science and perhaps I could’ve taken more stats and programming courses, but that wasn’t my focus then. 

Career paths, including mine, can be very non-linear. I think it was Steve Jobs who said that it’s easier to connect the dots when looking back, but harder to pave a path going forward. It doesn’t mean you shouldn’t plan. You absolutely should. It just means plans can change and that’s okay.

As a final thought: Don’t ever talk yourself out of going for an opportunity that you want to pursue. You find reasons to think you may not have all the qualifications, but it’s better to grow into a role than take something where you’re bored and not learning. Let someone else make a determination about whether you’re a fit. Rejection may sting a little bit, but it’s temporary. It’s better to take a shot on a goal with the possibility of missing than to not take a shot at all.