nanosamurai (nanosamur.ai)

nanosamurai is an open-source, production-grade speech AI platform designed for organizations that need to keep sensitive conversations private without sending data to third parties. It provides a model-agnostic orchestration stack supporting realtime, refined, and batch transcription with multitenancy, session persistence, speaker diarization, and a browser-based UI, API, SDK, and local observability tools. The platform is deployable locally with Docker Compose using various speech models and includes tools for recording, storage, and transcript auditing.