Show HN: I built an open source AI gateway that proxy calls to multiple LLMs

Clients can talk via HTTP or gRPC. Comes with rate limiting, failover, caching, guardrails, logging and an admin API for configuration

Love to get feedback if this is interesting to you