Introducing

livekit-memory

— RAG for LiveKit Agents Just shipped a new module for powering real-time contextual retrieval in your LiveKit voice agents. What it does: Chunks & indexes your knowledge sources into Qdrant Features: • Sub-20ms retrieval latency with localhost Qdrant which is real-time for voice agents • Language-aware code splitting (Python, JS, Go, Rust, etc.) • Semantic chunking with LLM or fast recursive splitting Installation:

pip install livekit-memory

Or,

poetry add livekit-memory

GitHub: https://github.com/joshiayush/memory PyPI: https://pypi.org/project/livekit-memory/ Right now it supports Qdrant as vector database, but contributions are always welcome to add other database options as well. Thanks