rag_ex

Elixir RAG library with multi-LLM routing (Gemini, Claude, OpenAI, Ollama), GraphRAG, knowledge graphs, modular retrievers/rerankers, composable pipelines, pgvector integration, advanced chunking, and tool-using agents. Fork of bitcrowd/rag.

Note: This is a fork of bitcrowd/rag. Credit to bitcrowd for the original implementation.

Features

Core LLM Capabilities

• Multi-LLM Provider Support: Gemini, Gemini, Codex (OpenAI-compatible), and Ollama
• Smart Routing: Fallback, round-robin, and specialist routing strategies
• Streaming Responses: Real-time streaming for supported providers

Modular RAG Architecture (v0.3.0)

• Retriever Behaviours: Pluggable retrieval with Semantic, FullText, Hybrid, and Graph implementations
• VectorStore Behaviours: Pluggable vector backends with pgvector implementation
• Reranker Behaviours: LLM-based and passthrough reranking
• Pipeline System: Composable RAG pipelines with parallel execution and caching

GraphRAG Support (v0.3.0)

• Entity Extraction: LLM-based entity and relationship extraction
• Knowledge Graph Storage: PostgreSQL-based graph with Entity, Edge, Community schemas
• Community Detection: Label propagation algorithm for entity clustering
• Graph Retrieval: Local, global, and hybrid graph search modes

Graph Storage Backends

rag_ex supports multiple graph storage backends:

PostgreSQL (Pgvector)

The default back