BrainTrust - Ship AI apps without the guesswork (ANKUR IS PRESENTING)

Learn how to ship LLM apps reliably. This talk covers evaluation tools, managing datasets, and prompt experimentation to avoid guesswork and ensure quality.

Overview

BrainTrust is a LLM development platform including a SDK and tools for evaluation, model graded evals, managing eval datasets, and a playground for experimenting with prompts.
BrainTrust is in production and used by multiple late stage startups already.
BrainTrust is founded by: Ankur Goyal (who recently exited a ML company to Figma) will come present the demo if accepted. [email protected]. Elad Gil helped incubate this company.

Links

https://braintrustdata.com/
Braintrust: An LLM evaluation platform for robust AI application optimization.
https://github.com/braintrustdata/autoevals
AutoEvals evaluates AI outputs via LLM-as-a-judge, heuristic, and statistical methods.

Tech stack