Jun 13Vibe with Hermes Agent — Bengaluru · RSVP
ToolsMCPBlogResearchCommunityStar on GitHub
All Tools
E
RAGFreeOpen Source

EDGEQUAKE

Rust-native GraphRAG — 10x faster ingestion, six retrieval modes

Apache-2.0

ABOUT

GraphRAG pipelines built in Python are slow — document ingestion and knowledge graph construction can take hours for large document sets, making iterative development impractical. EdgeQuake solves this by implementing the entire GraphRAG pipeline in Rust, achieving 10x faster ingestion than Python-based alternatives (LightRAG, etc.). It builds lineage-aware knowledge graphs that track document provenance, supports six retrieval modes (from simple keyword to complex graph traversal), and uses PostgreSQL for mature, production-grade storage. The result is a GraphRAG system that's fast enough for iterative development and robust enough for production.

INTEGRATION GUIDE

1. Ingest hundreds of documents into a knowledge graph in minutes instead of hours for rapid prototyping 2. Build a lineage-aware RAG system that traces each answer back to its source document and section 3. Deploy a production GraphRAG pipeline with PostgreSQL persistence, six retrieval modes, and high throughput 4. Power a knowledge discovery tool that lets users explore document relationships through graph traversal 5. Replace Python-based GraphRAG with a Rust-native alternative when ingestion speed is a bottleneck

TAGS

raggraphragknowledge-graphrusthigh-performanceretrievalopen-source
edgequake — AI Tool | Agentic AI For Good