Vibe Coding Discovermoss
View on GitHubThe retrieval layer for production AI systems. Lightning-fast (<10ms) search without vector databases. Built for browser, edge, on-device, and cloud.
Moss is an embedded semantic and hybrid search runtime for AI applications, with built-in embeddings, metadata filters, and browser/WebAssembly support. SDKs for Python and other languages support low-latency retrieval without a remote vector database on the query path.
Use Cases
Low-latency retrieval for conversational agentsVoice assistant knowledge searchIn-browser semantic searchRAG applicationsSearch over application documentsMetadata-filtered document retrieval
Built With
- Language
- Python
- Frameworks
- LangChain · DSPy · LlamaIndex · CrewAI · Haystack · AutoGen · Pydantic AI · Pipecat · LiveKit
Tags
semantic search · hybrid search · retrieval · RAG · embeddings · metadata filtering · on-device · WebAssembly · voice AI · real-time