Startup Europe — The Sifted Podcast Hertility CEO Helen O’Neill on building a foundational model for women’s health
Apr 30, 2026
Helen O’Neill, Associate Professor in reproductive and molecular genetics at UCL and CEO of Hertility, builds diagnostics and care pathways to improve fertility and gynecological outcomes. She talks about creating multimodal datasets and AI-driven diagnostics. She explains the technical hurdles of women’s health data, scaling clinician workflows with AI, and the fundraising biases around femtech.
AI Snips
Chapters
Transcript
Episode notes
Hertility Started As A Diagnostic Clinical Trial
- Hertility began as a clinical trial to build predictive diagnostic algorithms for gynaecological conditions linked to infertility.
- Helen launched the trial after a grant to study how insidious symptoms like pain and bleeding mask common conditions that delay diagnosis.
Why Women's Health Data Has Been Scarce
- Women's health data is scarce because female biology is 'hormonally noisy' and sampling must be timed to day three of the cycle.
- Helen O’Neill explains day three sampling gives only 12 reliable opportunities per year, making large-scale female data collection technically difficult.
Multimodal Labeled Data Enables High Confidence Diagnosis
- Building a multimodal, labeled dataset (health assessments, bloods, ultrasounds) enables high‑confidence AI diagnoses for women's conditions.
- Helen says their models now diagnose conditions with 98–99% confidence using this proprietary, annotated patient data.