Open-Source Version of Anthropic's Internal Analytics Engine

This open-source project replicates Anthropic’s internal analytics engine, providing developers and data teams with a self-hosted tool to track, visualize, and debug LLM interactions. It’s designed for engineers building AI applications who need granular insight into model behavior, latency, and costs without relying on proprietary services. The project is interesting because it democratizes access to production-grade observability typically reserved for large AI labs, enabling deeper optimization and transparency in custom LLM workflows.