7:19

AI x Bio Pitch Contest — Umbra Bio

Paul Donohoue presents Umbra Bio's approach to multi-attribute protein design and answers the judges' questions.

Xin

Chapters

  1. 00:00 Umbra Bio and the protein-design gap
  2. 00:52 Designing for multiple attributes
  3. 01:50 Wet-lab validation and team
  4. 03:57 Partner programs and next steps
  5. 04:59 Q&A: Customers and validation
  6. 06:21 Q&A: Long-term business model
Summary

Umbra Bio co-founder and CEO Paul Donohoue describes a protein-design platform that uses repeated prediction and scoring to consider more than binding to a target. He discusses the company's wet-lab validation, partner work, and plans to develop therapeutic assets. The judges ask about customers, structural inputs, and Umbra's long-term business model.

Watch next

More from the studio

TechCrunch Disrupt 2026 · AIEEV3:12Open Mic

TechCrunch Disrupt 2026 · AIEEV

AIEEV co-founder Hyeonkyeong (Karen) Kim explains the approach behind Air Cloud: connecting edge GPU devices for AI inference, supporting container-based deployment and autoscaling, and addressing hardware and network security. Recorded at TechCrunch Disrupt 2026.

Oct 22, 2026
AI x Bio Pitch Contest1:10

AI x Bio Pitch Contest

A short look at the September 29 AI x Bio Pitch Contest in San Francisco, from company pitches to the award presentations.

Oct 5, 2026
AI x Bio Pitch Contest — Optagon Labs9:25

AI x Bio Pitch Contest — Optagon Labs

Khalil Kaddoura presents Optagon Labs' proposed AI-driven loop for drug optimization and lab testing.

Oct 5, 2026