I've had PCOS for fifteen years. For fifteen years, I've struggled with weight and irregular periods, and for fifteen years every doctor told me some version of the same thing: eat healthy, lose weight, and the symptoms should improve.

I did what you're supposed to do. I saw primary care doctors, OB-GYNs, endocrinologists, nutritionists. We ran the standard tests, over and over: HbA1c, glucose, liver function, kidney function, complete blood panel. Everything came back normal. Every year. When I asked what was actually causing my PCOS, the answer was that PCOS doesn't really have a root cause — we can only manage symptoms, and the way to manage them is food and lifestyle.

But "food and lifestyle" without a mechanism is an impossible instruction. Social media says one thing, ChatGPT says another, every influencer has a protocol, and none of it agrees. PCOS care needs to be hyper-personalized, and I had nothing to personalize against. I was doing everything and couldn't tell if any of it was right.

The engineer's approach failed too

After I left Microsoft, I attacked the problem the way an engineer would. I took a Stanford course on sports nutrition and built myself an app — hourly guidance on hydration, macronutrients, and sleep, modeled as a supply-and-demand problem for the body. I followed it faithfully.

It didn't work. And the reason it didn't work taught me more than the course did: everything in sports nutrition assumes a baseline-healthy body. If a baseline-normal person cuts calories, they lose weight. I don't. Baseline-normal people don't gain weight seemingly at random. I do — and I couldn't say why. I was optimizing on top of a system whose actual state I couldn't observe. Any engineer knows how that ends.

By this point I wasn't just frustrated. I was starting to lose trust in my own body.

One blood draw

Five months later, I found a functional medicine doctor — outside the insurance system, which is its own indictment. She didn't tell me to eat healthy. She ordered tests I'd never been given in fifteen years of PCOS care, including a glucose tolerance test that measured my insulin — not just my glucose — at multiple points across two hours, with seven markers tracking how my body was actually responding.

The result: insulin resistance. My glucose looked fine because my pancreas was flooding my body with insulin to keep it fine. HbA1c can't see that. Fasting glucose can't see that. The standard panel had been reporting on the wrong layer of the system for fifteen years.

For the first time, I had a number. A measurable driver of my PCOS, a value I can track, and a mechanism I can work against. I'm now working with my doctor to lower my insulin resistance — and with it, reverse the symptoms I was told could only be "managed." The uncertainty that had defined fifteen years of my health collapsed into something an engineer can love: a metric.

The system is broken in a very specific way

Here's what I now understand. PCOS sits at the intersection of primary care, gynecology, and endocrinology — and those three specialists almost never sit at the same table for one patient. Each sees a slice. Nobody owns the system. And the tests that reveal root-cause drivers — fasting insulin, a full insulin curve, markers most women have never heard of — aren't part of the standard workup, even for women already diagnosed with PCOS.

Had I known about my insulin resistance five years ago, my life would look different. That sentence is why I'm building now.

What I'm building

Folia is where I've started: a daily companion for women navigating PCOS, with voice check-ins that become a structured diary, and a research chat that only answers from hand-curated papers, with citations. But the bigger arc is the one my own story traces:

  • Help women walk into their next appointment with a clear symptom history — and the vocabulary to ask for the right tests. More women should know to ask for fasting insulin or HOMA-IR by name.
  • Connect women with PCOS to doctors who actually look for root causes — ideally ones who take their insurance.
  • Eventually: make root-cause testing itself accessible — a single blood draw that tells a woman not just whether she has PCOS, but which subtype, which driver, and how severe it is right now.

If even one woman gets her root cause diagnosed five years earlier than I did, this will have been worth building.