This page may contain stale information. Last updated: 2026-09-19
Overview
Diogo Almeida co-created ChatGPT-era training methods including RLHF at OpenAI. Founded typesafe-ai (~2024) to build non-generative decision models; released jev September 2026.