Sasha Sheng is Co-Founder and COO at TypeSafe AI, a frontier AI lab pioneering machine-native, composable intelligence designed to run reliably inside software systems rather than interact with people.

Sasha studied Engineering at the University of Michigan and went on to become a research engineer at Meta/FAIR, where she worked across News Feed, AI Experiences, and AI Research.

An avid builder and community organizer, Sasha has published research at top-tier venues including NeurIPS and ECCV, and won 7 hackathons in 2023 alone.

Alongside co-founders Diogo Almeida (former OpenAI RLHF co-inventor) and Erik Gafni (repeat founder), Sasha helped build TypeSafe from stealth to a $40M seed round led by DCVC, with participation from AWS.

TypeSafe emerged from stealth in September 2026 with its flagship model Jev, delivering typed probabilistic decisions at 70ms-500ms, up to 200x faster than traditional LLMs, unlocking a new era of composable AI for developers.