A Python Search Engine for Humans 🥸

⚡ Cut Claude token usage by 90%+ — free, open-source, local-first context compression for Claude Code. Hybrid RAG (BM25 + ONNX vectors), AST chunking, reranking. No API needed.

Production-ready RAG system for technical documentation with hybrid retrieval, cross-encoder reranking, and RAGAS evaluation metrics.

Open-source framework for cold-start recommendations and lazy knowledge graph construction from product catalogs

🎵 专属你的私人数字调音师|AI 音乐搜索推荐 Agent | 基于大模型 + 知识图谱 + 双模型声学向量的本地智能音乐推荐系统 | LLM-powered Music Recommendation Agent with Hybrid RAG, Neo4j, and Long-term Memory

GraphRAG over aviation safety incident reports for causal reasoning and safety pattern extraction

Unified memory for all your A.I. agents. Knowledge graph hybrid retrieval of important context.

Tiny-RAGFlow is a lightweight Retrieval-Augmented Generation (RAG) framework designed for quickly building efficient vector retrieval systems.

Book discovery backend with hybrid retrieval, evidence-grounded QA, explainable recommendation, and circulation management.

Persistent memory for AI coding agents with outcome-based learning

Evidence-synthesis RAG assistant for TCM practitioners — hybrid vector + knowledge graph retrieval over 17 classical texts, with query classification, self-critique verification, and blind A/B arena evaluation.

Production-grade Node.js RAG system with hybrid retrieval, pluggable adapters, and OpenTelemetry tracing

VOTR: Vector Orchestrated Tool Retrieval for Scalable Multi-Agent Systems. Companion Paper:

Principal-grade technical reference for agentic AI systems: context engineering, MCP/gRPC/JSON-RPC, hybrid retrieval, memory architecture, tool orchestration, verification loops, and production reliability.

🧠 Universal long-term memory for AI agents. GraphRAG-powered knowledge base with vector search + graph traversal. Privacy-first, local-only, MCP-compatible. Connect Claude, Copilot, or any AI assistant.

Early 2026, constant attacks targeted AI-powered enterprise systems, staying on top of security news is critical. This repo demonstrates a pipeline to aggregate and triage IT security news from multiple sources. It uses hybrid retrieval (BM25 + semantic embeddings) with LLM-based reasoning, evaluated against a benchmark dataset.

Fully offline iPhone meeting memory. Zero network calls, no Wi Fi required. Records, transcribes, summarises, and answers questions entirely on device. Audio, transcripts, summaries, embeddings, and chat never leave the phone.

LumiCite is a multimodal RAG system for academic papers, designed for multimodal evidence retrieval and citation-aware question answering.

Hybrid resume–JD matching system using TF-IDF and Hugging Face MiniLM sentence embeddings with explainable hybrid scoring and deterministic feedback.

Baiyuan RAG Knowledge Platform Whitepaper — L1 Wiki + L2 RAG hybrid retrieval infrastructure powering Baiyuan's AI Customer Service, GEO, and PIF product lines. 3 languages (zh-TW/en/ja), 12 chapters + 4 appendices, CC BY-NC 4.0.