Meet the Team, Vol. 1

I didn’t think I’d be an entrepreneur. But, when I consider things in reverse, it makes sense. My parents came to the US from India in the 1970s, and I was born in Buffalo where I lived until I was six. My favorite memory is playing in the snow without the responsibility of having to shovel it.

I grew up in Phoenix, and early on I was interested in sports, math, music, and science. I played piano and saxophone in concert and jazz bands. During high school I took my first programming classes but mostly played Quake 3. Still, I learned computers are force multipliers.

When I was trying to pick a major at the University of Arizona in Tucson, a family friend suggested that healthcare and computers would be an interesting combo. Little did I know how that observation would play into my life! I majored in computer science and cell biology, but I lacked a clear sense of direction for my career. I had research experience (microarray data analysis), so applying to MD/PhD programs bought me time to figure out what I wanted to do (I had a better answer for my interviews).

Clinical informatics was off the beaten path when I started Vanderbilt University’s MD/PhD program. Teammates matter, and I was fortunate to meet my future PredictionHealth co-founders at the very beginning. I met Pedro Teixeira during our interview weekend and we were assigned to share a hotel room. I met Michael Poku during my housing search and we lived in the same building.  

The 2009 healthcare reform debate during my first year at Vanderbilt brought challenges like quality and value of care into focus. I saw fragmented care delivery across clinical teams that were patched together by inefficient workarounds. In the Department of Biomedical Informatics, I learned how messy data slows discovery and the importance of fitting into the real time clinical workflow.

Pedro and I discussed the challenges in healthcare and how to improve wellness and care quality. In 2013, the American Medical Informatics Association held their first Student Design Challenge which focused on reinventing clinical documentation. While our system is very different today, our goal then was the same – to support providers in delivering the best patient care. We finished 2nd in the challenge (too business-y and not academic enough). Afterward, we thought, “let’s just start building it,” and so that’s how we spent our Saturdays. We took what was supposed to be a short leave of absence near the end medical school to build a product and launch it. 

We wrote and then deleted hundreds of thousands of lines of code. Fortunately, we learned that figuring out how to solve your users’ problems was the critical ingredient for a successful company. We also learned that we couldn’t stop building tools to help clinicians – applying many years of clinical, research, and computational experience to make patient care better. Today we’re growing a startup called PredictionHealth with customers all over the country, and we’ve learned many valuable lessons along the way:

• Launch your first version quickly and iterate until you provide value
• Listen to your users!
• Make data-driven decisions with metrics
• Build something your customers truly love

I’m excited about building the data infrastructure that healthcare is missing to make incorporating AI into the clinical workflow possible. Google and Amazon built cloud infrastructure that helped companies grow and innovate more quickly. Similarly, our infrastructure will make it easier to train, evaluate, and incorporate new AI models into clinical workflows. We can make healthcare better for every patient and we can do it faster with the right tools.

As for personal goals, I’m just glad I finally got through the gauntlet of the last years of medical school while building a company. I think things should get easier from here – at least I hope so!