CocoIndex
Incremental data indexing framework that keeps AI agent context fresh
What it is
CocoIndex is a Python framework with a Rust core for building data pipelines that turn codebases, documents, PDFs, chat logs, and other sources into indexes for AI agents and LLM apps. Pipelines are declared in Python, and when source data changes only the affected parts are reprocessed, so vector stores and knowledge graphs stay current without full rebuilds. Developers building RAG systems, semantic search, or agent memory use it to keep retrieval data in sync with live sources.