TypeSafe AI

TypeSafe AI is an AI lab founded in 2024 in San Francisco by Diogo Almeida, a former OpenAI researcher and co-author of the GPT-4 paper, together with Erik Gafni and Sasha Sheng. The company builds machine-native intelligence infrastructure for software automation and introduced the System One Model category; its first model, Jev, does not generate text but classifies inputs into predefined outputs with calibrated probabilities, targeting high-frequency, low-latency structured decisions inside software. Jev opened for early access on September 15, 2026, alongside a US$40 million seed round led by DCVC at a post-money valuation of about US$200 million; the company is reported to be in talks for a new round of at least US$1 billion at an indicated valuation above US$10 billion (not yet closed).