Pezzo

A self-hosted prompt management platform with monitoring and caching for LLM applications, plus Node.js, Python and LangChain clients.

Pezzo helps developers and teams manage prompts and investigate problems in LLM applications. It brings prompt editing and monitoring into one place for projects where several people contribute to AI behavior.

Prompt management lets teams design prompts, track versions and deliver changes to their applications.
Collaboration gives contributors a shared place to maintain prompts.
Monitoring and troubleshooting help developers inspect AI operations and investigate issues.
Caching can reduce response latency and model costs, with support across the Node.js, Python and LangChain clients.

The full platform can run locally through Docker Compose, so you can host its prompt management and monitoring services on your own infrastructure. Pezzo is open source under the Apache 2.0 license. Its infrastructure uses PostgreSQL, ClickHouse, Redis and Supertokens.

LLM Observability and Tracing Prompt Management and Playgrounds Docker Web Multi-user access Prompt versioning