Skip to content

Heather Bowerman, CEO and Founder of DotLab


Heather Bowerman is the CEO of DotLab, a personalized medical company which developed the world’s first non-invasive test for endometriosis, a condition affecting 10 percent of women. Before creating DotLab, she was a consultant in healthcare and technology for McKinsey & Company and later led business operations at Enlitic, a trailblazer in using artificial intelligence to help doctors better interpret radiologic images for diagnoses. She also served as a policy associate in the White House’s Office of Science & Technology Policy during the Obama administration.

Ms. Bowerman has received countless honors for her work in health technology, including being recognized as one of the world’s top 35 innovators by the MIT Technology Review; as one of the 100 most intriguing entrepreneurs by Goldman Sachs; and as a finalist for the World Technology Award for Health & Medicine. She also placed among the Forbes list of top 10 female-founded startups and received the American Society for Reproductive Medicine Prize for Scientific Research. She holds a bachelor’s degree in bioengineering from the University of California, Berkeley, and was a teaching and research fellow at Harvard University.

Ms. Bowerman graciously gave a 30-minute interview, which has been edited for length.

Interview Questions

[] How did you initially get into artificial intelligence, and specifically into deep learning?

[Heather Bowerman] I knew that artificial intelligence was this great way to scale the impact that I wanted to make in medicine and technology. When the press talks about AI, what they’re often referring to is deep learning, which is a particular family of algorithm.

When I was thinking about career moves, I was really interested in a company working on this called Enlitic, which was the world’s first ever medical deep learning company. I joined as the fourth employee as the senior researcher and later led business operations.

As a little bit of background, deep learning involves training artificial networks on large quantities of data such as medical images, whether it be CT scans, MRIs etc.—and then getting them to provide inferences about new data. Meanwhile you have radiology, which is one of the most expensive operational items for healthcare organizations. Radiologists are involved in all aspects of the continuum of care, but there’s underfunding and staffing shortages. I was interested in solving the backlog via AI, since it negatively impacts diagnosis and access to the appropriate treatment for patients. Radiology within healthcare was such an exciting application for deep learning.

[] How did your work at Enlitic inform your founding of DotLab?

[Heather Bowerman] In all my prior roles, including Enlitic, I saw how research and the deployment of technology severely lagged in medicine, but particularly in women’s health, which led to worse outcomes for women. I started thinking about what the best point of entry could be to eliminate the systemic bias and bring healthcare quality to women. There’s a really great report from the Brigham & Women’s Hospital in Boston that describes how healthcare is optimized for men. Just seeing the opportunity for impact was the biggest thing for me.

[] You’ve worked at the White House, and you’re one of the Goldman Sachs 100 most Intriguing entrepreneurs. You’ve gotten accolades from the MIT Technology Review, among many others. Through all of your roles and accomplishments, who have been your greatest mentors?

[Heather Bowerman] I love that question. My greatest mentor has definitely been my mom, but besides her, the mentors I’ve learned the most from have been in academia. I was really inspired by Marian Diamond, who was a professor at Cal and passed away recently. Many of my professors have had a great impact on me.

I also read as much as I can. I love learning about leaders in other industries, and I think there are always opportunities to learn from people of all stripes. I learn from kids I mentor who are just coming out of high school. Also, I spent a whole day at the Apartheid Museum in South Africa this year and learned about the impact of Nelson Mandela’s leadership. You can draw inspiration from many different sources.

[] I agree. I want to talk about the demographics of bioengineering and AI. Would you say that in most business situations you are one of the only women in the room or is it more gender equitable?

[Heather Bowerman] My company DotLab has 50/50 gender parity right now, but going back to the days of being a student, women made up about a quarter of the undergraduate student body in engineering at Berkeley. Some departments were higher than others. I’ve seen a report that says that Cal engineering graduation rates are indistinguishable by gender, and I think this is partly due to the administration and the faculty at Berkeley where women hold two in five associate dean positions in the college and around 30 faculty positions.

Organizations where I’ve worked have usually skewed male and sometimes dramatically so. There’s a prominent leader in AI at Google, Jeff Dean, and he said that he was really worried about the lack of diversity in AI. But if you look at the Google Brain team, just scrolling through the photos it’s something like 95 percent male and I see just three women on there! It’s also over 70 percent white. Another prominent group, OpenAI, doesn’t share their demographics at all.

[] Yes. Most big tech companies, particularly in their engineering departments, seem to skew heavily male. What do you think would be some of the negative outcomes of having AI products designed predominantly by white men?

[Heather Bowerman] Heavy question. There are a lot. The gender diversity piece is a really important component, but let’s talk about another example. There are algorithms that are affecting the situation in Myanmar right now, where many people have Facebook as their only news source. Misinformation and propaganda are being spread about the Rohingya, an ethnic minority group there, and there have been unspeakable tragedies as a result.

I think with diverse teams, problems like this could be minimized—to help really think through the consequences of what we’re building.

[] That’s a great answer. I’m trying to visualize what it would look like in terms of a funneling or a bottlenecking of one group’s truth that doesn’t necessarily reflect the facts. Am I understanding you?

[Heather Bowerman] Think of it this way: diversity could help minimize the runaway feedback loop that we see when there’s not a diverse team in place. It’s really just applying ethics to a given topic.

For example, there’s, a website that most people know. I watched this YouTube video and an employee presented evidence that the site’s algorithm suggested text meetups to women less frequently than men. This runaway feedback loop resulted in fewer females learning that these meetups were happening, and then in turn, the algorithm would recommend fewer meetups to those girls and women. It’s a self-fulfilling prophecy.

In the video, the man from Meetup talked about how his team made an ethical decision for their algorithm to not create such a feedback loop because of the consequences it would have on the the women who would learn about the meetups. It’s encouraging to see leadership to avoid that outcome.

Going back to the example of the situation in Myanmar: it’s a narrative that could use an ethical intervention instead of just letting the algorithm run wild.

[] That’s really a tough set of issues to address. You had mentioned earlier that there is a lack of women working in tech and AI specifically. Why do you think that women are still underrepresented in these roles?

[Heather Bowerman] Some of the best work I’ve seen on that has been from the McKinsey report where they break out technology to examine that question. It’s hard to say exactly why. It’s a multifactor situation.

I do think that things are improving. The adage that sunlight is the best disinfectant is something that we’re seeing. I hope that eventually the transparency and pressure on companies to report their gender diversity statistics will help to drive change.

There’s also a shift toward support for mission-driven technology companies, including topics that might appeal to women uniquely. I see my company DotLabs as an example of that, where women might be more excited to work for companies focused on problems that are meaningful to them. We’re working to create technology that can help women live more productive, healthy lives, and that really is the underlying motivation for everything that we do.

[] That’s really great to hear. Apart from radiological diagnostic tools, what are some of the other applications of AI in healthcare that people can consider?

[Heather Bowerman] People really got excited about applying deep learning to healthcare for the first time during a competition that Merck sponsored in 2012. Participants were given a dataset describing the structure of different molecules. Their challenge was to build an algorithm that would predict which of those would be the most compelling candidates for drugs. The team that won that competition entered at the last minute and had no domain knowledge of biochemistry. That team used deep learning and won the entire competition.

[] So they were computer scientists or statisticians? They had nothing to do with healthcare?

[Heather Bowerman] They were academics who happened to focus on deep learning!

[] Wow, that’s incredible.

[Heather Bowerman] I’m also really excited about the applications of deep learning to emerging economies. They often have fewer barriers to dramatic change compared to ones which are more developed. In other words, there’s less friction when you’re coming into an emerging economy as a startup to deploy your technology.

I hope that the new generation of entrepreneurs develops solutions that can identify the best health system models—ones which avoid the problems seen in developed Western countries. I think there’s real opportunity with deep learning in healthcare to build sustainable healthcare systems in those areas.

To give a specific example from an economic forum report: they’re saying that in Nigeria, the number of positions per capita would have to increase by 12 times between now and 2030, requiring 10 times the current annual public health spending. Basically, it would take 300 years with the existing training structure for a position in order to catch up with OECD economies.

Bringing in AI is this amazing opportunity to have a positive, transformative impact on the doctor shortage there. There’s a choice to deploy the long, expensive and unsustainable path of developed economies to set up new health systems—or we can bring in the latest and greatest and really learn from mistakes that we’ve made in other geographies

[] It’s exciting just thinking about the opportunities to scale and adopt health systems that the world hasn’t even seen yet.

[Heather Bowerman] There’s just so much opportunity. I can speak mostly about healthcare, but really you can apply it to any industry.

[] Absolutely. You’ve already exacted so much positive change through all of your work. My last question is what advice do you have for women who are interested in learning more about AI and getting into the industry?

[Heather Bowerman] My biggest piece of advice would be to learn how to code. It doesn’t need to be your major in college. Or if you’re out of school: just getting your hands on whatever materials you can. There are some blogs with great resources for free courses to get the introduction to deep learning and AI.

Also, know your worth in these conversations. We have a lot of white men in AI and we know that getting deep learning into the hands of a more diverse group—women and people of color—can help bring a new point of view and new solutions to problems.

Last, think about what problems excite you and find out about companies that are working on them. What’s really helped me in my career is finding warm introductions to leaders in the field.