Bantanium

I started Bantanium because I wanted to answer a simple question:

How does my body actually work?

Living with PCOS/PMOS made me realize how difficult that question is to answer. Despite unprecedented access to health data, most advice is still remarkably generic. Two people can eat the same meal, follow the same training plan, or take the same medication and experience completely different outcomes. Yet our tools rarely help us understand why.

What began as a personal question has grown into a broader one: What would it take to build a system that truly understands an individual over time?

Why the PMOS Renaming Matters

The transition from Polycystic Ovary Syndrome (PCOS) to Polyendocrine Metabolic Ovarian Syndrome (PMOS) is more than a name change — it's an acknowledgment that we've misunderstood the disease for decades. The old name emphasized ovarian cysts, even though many patients don't have them, while overlooking the metabolic and endocrine dysfunction at the core of the condition.

To me, the significance isn't the terminology itself. It's that an entire field has agreed the original framing no longer describes the problem. Medicine is beginning to recognize PMOS as a complex, lifelong metabolic condition, but most consumer health products still reflect the old mental model: disconnected symptom trackers, calorie counters, and generic advice.

That gap is where I see opportunity. The science has evolved faster than the software. If our understanding of the disease has fundamentally changed, then perhaps the tools should change too — not by offering more information, but by helping each person build a personalized understanding of their own physiology over time.

Today's health ecosystem is fragmented. Wearables measure movement and sleep. Nutrition apps track food. Labs capture biomarkers. Clinicians see snapshots during appointments. Each contributes a piece of the story, but none maintains an evolving understanding of the person behind the data.

I don't believe we're missing data.

I believe we're missing understanding.

No clinician — no matter how skilled or attentive — has the time or memory to continuously synthesize years of sleep, nutrition, symptoms, biomarkers, and movement into an evolving model of a single patient. I believe AI may finally make that possible.

Healthcare doesn't need another chatbot that summarizes yesterday's metrics. It needs systems that learn alongside us — forming hypotheses, recognizing patterns, updating beliefs as new evidence arrives, and making uncertainty explicit rather than hiding it.

That's what Bantanium exists to explore.

Questions I'm Exploring

Rather than beginning with a product, I'm beginning with questions.

• What signals genuinely matter over months and years?
• Which relationships are causal, and which are coincidence?
• How should a system update its understanding when new evidence contradicts previous assumptions?
• How can uncertainty become something a system communicates instead of conceals?
• What does personalized health look like when it learns continuously instead of starting over at every appointment?

These questions matter more to me than any particular implementation.

Current Experiments

Bantanium is my research environment for exploring those questions.

I'm collecting and integrating longitudinal data from wearables, nutrition logs, sleep records, symptoms, biomarkers, and continuous glucose monitoring. Alongside that, I'm studying existing health products, reading scientific literature, and building small prototypes to understand where current approaches succeed — and where they fall short.

I want to earn the right to build.

The goal isn't to prove a preconceived idea. It's to become increasingly difficult to fool myself.

Principles

Looking Ahead