adaptive-rag

Training code for advanced RAG techniques - Adaptive-RAG, Corrective RAG, RQ-RAG, Self-RAG, Agentic RAG, and ReZero. Reproduces paper methodologies to fine-tune LLMs via SFT and GRPO for adaptive retrieval, corrective evaluation, query refinement, self-reflection, and agentic search behaviors.

Agentic Adaptive RAG is a production-ready framework for building self-correcting, reasoning-based LLM systems that dynamically choose between retrieval, web search, and generation.

A chat-based AI travel planning assistant that helps users create personalized travel itineraries using real-time data from TripAdvisor

Self-correcting agentic RAG with LangGraph — combines Adaptive, Corrective, and Self-RAG. Routes queries, grades retrieved documents, and checks answers for hallucinations. Runs fully local on Ollama.

"A resilient RAG agent built with LangGraph and LlamaIndex that self-corrects by rewriting queries when retrieval quality is low. Features a cyclical graph architecture for adaptive search.

An Adaptive-RAG system for course selection, featuring intent-aware retrieval, hybrid search, conversational web interface, and a comprehensive benchmark.

adaptive rag, corrective rag and agentic rag examples using langgraph

Adaptive RAG (Web search + Corrective RAG) is an upgraded model of Agentic RAG

Adaptive RAG pipeline with a lightweight DistilBERT query router for dynamic, latency-optimized retrieval strategies.

AI-powered educational assessment platform using RAG, LangGraph multi-agent workflows, FastAPI, React, Expo, and PostgreSQL for grounded exam and quiz generation.

A self-correcting RAG agent in LangGraph — routes questions, grades its own retrieval, falls back to web search, and refuses answers that aren't grounded in its sources.

Adaptive RAG is an advanced retrieval-augmented generation system that intelligently combines dynamic query analysis with self-corrective mechanisms to choose the most effective strategy for answering user queries.

Deep Agentsライブラリのアーキテクチャを活用した、ハーネス構成のRAGシステムです。複数のエージェントが協調動作し、質問の粒度に応じて最適なRAGシステム（Naive RAG / ColBERT RAG）を自動選択します。

End-to-end Adaptive RAG system with JWT authentication. Uses LangGraph to route queries between document retrieval (FAISS), general LLM knowledge, or live web search (Tavily). FastAPI backend, Streamlit UI, Gemini embeddings + LLaMA 3.3 for generation.

Feedback-Driven Adaptive RAG System using LangChain, Groq, TF-IDF, and Streamlit for accurate, context-grounded question answering with hallucination prevention.

Local-only, graph-augmented, iteratively-reasoning, self-corrective RAG showcase on Apple Silicon — pushing toward published SOTA on multi-hop QA (MuSiQue, 2WikiMultiHopQA, HotpotQA) with HippoRAG 2 + Adaptive routing + Search-o1 iterative reasoning + CRAG-style gating + NLI faithfulness verification.

Context-aware tool for automated BDD test generation and execution using RAG, VectorDB, and LLaMA.

🏥 Modular Agentic RAG Medical Assistant 🤖 built with LangGraph, Pinecone, Groq & Tavily. It intelligently routes queries between RAG, live web search, and direct answers. 🔬 Live tracing shows every agent step, while FastAPI + Streamlit deliver a production-ready full-stack experience. 🚀

Enhanced adaptive RAG system with multi-turn reasoning and reflective query decomposition.

An attempt to production-ready Retrieval-Augmented Generation (RAG) system with advanced features including hybrid retrieval, adaptive feedback loops, comprehensive evaluation, and explainable AI logging.