How Numan used LangGraph and LangSmith to build Nu, an evaluation-first AI health coach.
UK digital health company Numan reviews its AI health coach Nu's evaluation system. The key finding is that scarce assets are not prompts but evaluation standards. Single-turn static datasets and scripted multi-turn conversations fail to cover real interactions. Nu now uses category-based scoring topic datasets, employs cheaper models to simulate patient dialogue, and lets safety checks automatically block releases.
Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open AIHOT original