Phenix

AI phenotyping platform for personalized PCOS care

Product Manager

2026

Visit Website
1 in 10
Women affected by PCOS
4
Metabolic phenotypes modeled
5
Core product modules shipped

The problem

Most PCOS tools treat all users the same, despite major metabolic differences across patients, leading to weak engagement and low information quality.

My process

Designed phenotype-based onboarding, an OCR food scanner with AI metabolic reasoning, a non-linear cycle tracker for irregular patterns, and evidence-backed treatment guidance.

The outcome

Shipped an end-to-end health product focused on root-cause care with personalized recommendations, clinician-ready summaries, and a scalable AI-driven architecture.

Key decisions

  • Positioned Phenix as a decision-support system rather than a generic tracker
  • Personalized every recommendation by phenotype to improve relevance and trust
  • Added doctor-export summaries so users can advocate with data in clinical visits