Bryan Smith has worn a lot of hats: data scientist for Team USA, three-time startup founder, venture partner. But none of them prepared him for what he saw during the postpartum period—until he lived it. With four kids and a wife who works as a high-risk postpartum nurse, he saw firsthand what the system was missing. Then he built Phia to fix it.

We sat down with Bryan to talk about what AI can actually do for postpartum care, why no one else tackled this window, and how personal chaos became a blueprint for one of the most clinically grounded maternal health platforms in the U.S.

Q: You weren’t a healthcare guy at the start. What pulled you into postpartum care?

A: My wife works in MFM and postpartum nursing, so I got a front-row seat to the gaps in care. We’ve had four kids—some smooth, some scary—and every time, I noticed how unsupported the recovery window was. From discharge to 3-4 months, moms are basically flying blind. There’s no infrastructure. No coordination. I kept thinking: if this were a different industry—sports, finance, tech—we’d call it negligence.

Then I saw a much deeper problem: it wasn’t just inefficient. It was dangerous.

Q: Where did the idea for Phia actually come from?

A: Watching my wife triage calls at night. Seeing moms spiral after birth. Seeing mortality rates that just shouldn’t exist in a developed country. And realizing AI could help—but not in the “chatbot replaces doctor” way people assume.

I’d worked in predictive modeling. I knew how to build systems that respond in real time, learn from signals, escalate when risk appears. But nobody was applying that to postpartum care. I saw an opportunity to catch warning signs—mood changes, pain patterns, blood pressure spikes—before they became ER visits or worse.

That’s where Phia started. Not as a product. As a response.

Q: Why is the 120-day window so critical in your view?

A: Because it’s the most ignored and the most dangerous. We act like childbirth ends when the baby’s out. But mothers die after. Babies get readmitted after. Families fall apart after.

From a data standpoint, the first four months are high-risk, high-noise, and completely unstructured. That’s exactly the kind of space where AI + operations design can outperform the current system. You don’t need 20 visits. You need the right care triggered at the right moment—before the cliff.

Q: What’s different about how Phia uses AI?

A: We don’t automate care. We automate attention. Every message, every BP reading, every symptom gets read and triaged instantly. Low-risk stuff gets handled fast. High-risk stuff gets escalated to a nurse or coordinator immediately. And over time, the system learns what patterns predict trouble for each individual.

It’s like having a nurse in your pocket who never sleeps—and a coordinator who sees your whole story, not just what you remembered to mention.

Q: You’ve built startups before. Why this one? Why now?

A: Because this is the one that matters most. If you want to change the next generation of health outcomes, you start with mothers. You start with that first 120 days. It affects everything—mental health, child development, economic mobility, mortality.

This isn’t a niche. This is foundational public health, and nobody was treating it that way. We are.

Q: What do you want Phia to stand for five years from now?

A: That we gave moms and babies a real system. That we lowered mortality with precision, not bureaucracy. That we built care that scales without losing its soul. And that every parent who went through our program felt held—medically, emotionally, functionally.

Because right now, most of them don’t.

Final note from the team:

Bryan’s vision shaped Phia’s core from the start: use the precision of AI to bring humanity back to postpartum care. His experience—as a data scientist, a founder, a partner to a postpartum nurse, and a dad of four—grounded our mission in both logic and love. Phia was built to close the gap that almost every parent walks through blindly. And we’re just getting started.