Braintrust LLM Evals Platform

My LLM evals/o11y platform of choice is Braintrust.dev.

There are a lot of LLM observability & eval platforms: Langfuse, Langsmith, Arize, Braintrust and Humanloop to name a few. Every existing player has added LLM observability capability.

Braintrust isn't the most established, fastest or most cutting edge. Where it shines is a focus on bringing Product + Design + Engineering together to surface and curate tests, datasets, evals, and experiments.

Zapier's setup walkthrough from AI Engineer conf: https://www.youtube.com/watch?v=fOtTHWeU6B8