An open-source LLM engineering platform for tracing what an application's prompts and agent steps actually did in production, plus tools for prompt management and running evaluations against test datasets.
Pricing: Open-source