AlleyCorp led the $27.6M round. Uncork Capital, Renegade Partners, SemperVirens and Jack Altman also took part. Epsilon had stayed quiet about the raise until now.
The San Francisco company took an unusual route. Rather than sell a model for one disease or modality, it became the radiology practice itself. Radiologists stay in the loop. The AI was built in-house for the practice’s own clinicians, not licensed out to other groups.
Volume has followed. The company has served more than 250,000 patients in under 10 months. It now handles over 2,500 imaging studies each day. By its own forecast it will read 1% of all daily US X-rays during 2026. One large outpatient imaging provider already sends it more than half its volume.
Radiology underpins an estimated 75% of medical decisions, and the radiologist workforce is not keeping pace with demand. That gap is Epsilon’s opening.
Leadership spans both fields. Its head of machine learning once led research at Google DeepMind. Its chief medical officer previously held that role, and the chief of strategy post, at Envision Radiology, a group of more than 500 radiologists.
The money funds market expansion, hiring, clinical partnerships and infrastructure. Owning the practice hands Epsilon something model vendors lack, a continuously refreshed stream of real-world cases that loops back into training.
