aigc-apps/PAI-RAG

An easy-to-use framework for modular RAG

What it solves

PAI-RAG is an open-source Agentic RAG framework designed for enterprise-level complex knowledge processing and intelligent Q&A. It moves beyond traditional RAG by integrating knowledge base management, multimodal understanding, and an agentic reasoning loop to handle complex tasks that simple retrieval cannot solve.

How it works

The system uses a ReAct (Reasoning and Acting) agent engine that follows a "think-act-observe" loop, allowing it to plan tasks and call tools in parallel. It combines multiple retrieval strategies—including vector, full-text, and hybrid search with reranking—to fetch information from knowledge bases supporting over 20 document formats. It also integrates the Model Context Protocol (MCP) to extend its toolset and supports multimodal inputs (images and video) by extracting descriptions and indexing them.

Who it’s for

It is built for developers and enterprises who need to build sophisticated AI agents capable of data analysis, multimodal document understanding, and real-time web searching without requiring deep machine learning expertise.

Highlights

  • Multimodal Capabilities: Extracts and indexes images from documents and supports video parsing to generate descriptions.
  • Agentic Reasoning: Native support for Chain-of-Thought models (like DeepSeek-R1) and task planning for multi-step queries.
  • Enterprise-Ready: Features multi-tenancy, RBAC (Role-Based Access Control) down to the chunk level, and OpenTelemetry for observability.
  • Extensible Tooling: Includes a built-in Python code sandbox, Text-to-SQL for database queries, and deep MCP integration for external tool access.
  • Broad Compatibility: Supports seven different vector database backends and any OpenAI-compatible LLM.

Related

  • Project
    langflow-ai/openrag

    An intelligent agent-powered document search platform that enables users to upload, process, and query documents through a chat interface using RAG.

  • Project
    java-up-up/nexus-agent

    An enterprise-grade AI agent platform that implements a complete loop from document governance and hybrid RAG retrieval to ReAct-based agentic reasoning and MCP tool integration.

  • Project
    nageoffer/ragent

    Ragent AI is a production‑ready Java platform for building Agentic RAG applications. It provides hybrid retrieval (vector, keyword, graph, web), advanced query understanding, model tiering with fault‑tolerance, multi‑level session memory, Redis‑based traffic protection, and a full ingestion‑to‑answer loop. The code is split into seven Maven modules (infra‑ai, system, rag, agent, bootstrap, framework, mcp‑server) that isolate AI vendor, storage, and retrieval concerns, making components interchangeable. v2 uses a ReAct‑style agent loop with MCP tool calls, Skills, and human‑in‑the‑loop confirmations. Production features include circuit‑breaker, transactional RocketMQ, idempotency, tracing (Langfuse optional), SSE streaming, and Sa‑Token security. A React UI offers both chat with citation/agent view and an admin console for knowledge‑base and intent‑tree management. Targeted at Java backend engineers and interview‑preparation candidates who need a real‑world, extensible AI stack rather than a toy demo.

  • Project
    GiovanniPasq/agentic-rag-for-dummies

    The repository provides a concrete, modular example of an Agentic Retrieval‑Augmented Generation (RAG) system built with LangGraph. It covers hierarchical document indexing, multi‑agent query processing, human‑in‑the‑loop clarification, and supports multiple LLM providers (default Ollama). Includes notebooks for setup, evaluation, and observability, making it a solid learning and prototyping resource.

  • Project
    erthorpabar/pabar_agent_runtime

    An AI agent runtime framework that integrates chat, RAG, and agentic loops to enable professional automation tasks across desktop and mobile platforms.