What connects it

On the surface, my career looks like a series of stops: academia, then health tech, then a consumer platform, then another. I’ve also run six marathons and spent time as a nutrition and habits coach. It can look scattered if you don’t know what connects it.

What connects it is one question I’ve been circling my entire career: how do we help people live longer, better lives — and what does data have to do with it?

Here’s the personal version of that question.

When I first moved to the US, I was training for marathons and constantly exhausted in a way that didn’t make sense. I kept asking my doctor for a ferritin test — I had a hunch something was off with my iron. I was told, repeatedly, that I didn’t need one. My hemoglobin was barely normal, and that was good enough.

It wasn’t until I started working at InsideTracker and finally got my ferritin tested that I found out it was at 10. To put that in context: it’s almost miraculous I could walk up a flight of stairs, let alone run a marathon.

I’ll never know for how long I’d been iron-depleted, or what I might have done differently if I’d known sooner. I can’t count how many times I was told to just run less. What was missing was a doctor with the time and interest to look beyond “she’s young and healthy.”

We’re at a strange moment. People have more access to health data than ever — wearables, continuous glucose monitors, detailed bloodwork, DNA tests. The data exists. The curiosity exists. And yet most people have no idea what to do with any of it. A doctor’s visit ends with “your numbers look fine,” because most physicians aren’t trained to engage with this kind of data, and frankly don’t have time to. People are left alone with dashboards they can’t interpret, making decisions based on guesswork.

I’ve spent my career — from computational biology research at Harvard Medical School and Dana-Farber, through building biological age models at InsideTracker, to large-scale ML at Zwift, Nuna, and Meta — trying to close that gap. Not just by building better models, but by asking what it actually takes for data to change how someone understands their own body and what they do about it.

I know what it feels like to have the data dismissed. I also know what it feels like when the right number finally surfaces and everything clicks into place.

That’s the work. And I’m not done with it.