apogee-ai-rag¶
Production RAG toolkit built on Clean Architecture. Compose any of 28 RAG variants over 60+ integrations: vector databases, embedders, rerankers, chunkers, parsers, evaluators and observability tools.
Highlights¶
- 28 RAG variants — Naive, Advanced, Modular, Self-RAG, CRAG, Adaptive, Agentic, GraphRAG, KAG, CAG, HyDE, RAG-Fusion, RAPTOR, FLARE, ReAct, Hybrid, Conversational, Multi-Modal, Long Context, …
- 60+ integrations — pgvector, Qdrant, Weaviate, Pinecone, Milvus, FAISS, Chroma, Elasticsearch, OpenAI/HF embedders, BGE/Cohere rerankers, LangChain/LlamaIndex bridges, Ragas/TruLens evaluators, Phoenix/Langfuse observability.
- Zero API keys to start — fully in-memory naive pipeline runs offline.
Install¶
pip install apogee-ai-rag
pip install "apogee-ai-rag[pgvector]"
pip install "apogee-ai-rag[qdrant]"
pip install "apogee-ai-rag[langchain]"
pip install "apogee-ai-rag[ragas]"
pip install "apogee-ai-rag[all]"
Read next¶
- Installation
- Quickstart — naive RAG in 20 lines
- Variants — when to pick which RAG type
- Integrations — vector DBs, embedders, rerankers
- Reference