Your local RAG stack — no APIs, no cloud, full control.

Key Features:

• Intent-based adaptive RAG with self-correcting retrieval (auto-retry on validation failure)
• Ensemble search via Reciprocal Rank Fusion combining BM25 + semantic vectors
• Local-only: Ollama for embeddings/LLM, Chroma for vector store, no external APIs
• Analyze programming languages via Docker Compose demo stack

# 1. Start Chroma vector database
docker compose up -d

# 2. Ingest programming language and conceptual documents
ragchain ingest

# 3. Search ingested documents
ragchain search "functional programming paradigm" --k 4
ragchain search "memory management" --k 5

# 4. Ask questions with RAG + LLM
ragchain ask "What is Python used for?"
ragchain ask "Compare Go and Rust for systems programming"
ragchain ask "What are the top 10 most popular languages?"

# 5. Evaluate RAG quality with LLM-as-judge
ragchain evaluate

# Clean up
docker compose down -v

Requirements:

• Docker (for Chroma)
• Ollama with bge-m3 (embeddings) and deepseek-r1 (generation)
• Python 3.12+

The ragchain ask command adapts to query type:

See AGENTS.md for architecture.