Philip Tannor

Philip Tannor is co-founder and CEO of Deepchecks, an open-source-first platform for continuous validation of machine learning and LLM systems. With Shir Chorev he raised a $14M seed to turn model testing from a notebook ritual into something that runs with the rest of the stack.

Open source for model testing

Deepchecks began as tooling practitioners could run to check data integrity, model performance, and drift. A TechCrunch seed piece framed the company around continuous validation rather than one-off evaluation scripts.

Seed and the CI instinct

Alpha Wave Ventures led the seed with Israeli investors including Hetz and Grove. Tannor's conference appearances (including Lynx events) present Deepchecks as infrastructure that sits beside CI for teams shipping models, not a consulting overlay.

On making ML failures visible

He talks about the cost of silent model failures in production and the need for tests that non-research engineers can own. That plain ops language is the through-line from open-source packages to the commercial platform.