#bm25

MCP-native code retrieval for AI agents — 84-88% fewer read tokens, BM25F + semantic search, AST chunks, session dedup

AI agent knowledge & long-term memory · MCP server, knowledge graph, BM25 + RRF retrieval · drop-in for Claude Code/ Openclaw/ Cursor / Codex / Hermes

AI document assistant with a production RAG pipeline — hybrid BM25 + semantic search, cross-encoder reranking, and streaming answers. Built with FastAPI, Streamlit, ChromaDB, and DeepSeek LLM. Fully Dockerized and deployed on Railway.

Agent + Multimodal RAG 全栈学习项目：LangGraph PAE/ReAct、Multi-Query、BM25 + pgvector 混合检索、RAG-Fusion/RRF、MinerU/OCR/ASR、evidence 引用与 HITL；React + Spring Boot + FastAPI。

Nokori (残り) — behavioral memory layer for Claude Code and Cursor. Turns your corrections into rules that come back when it counts. 经验留下的痕迹，比记忆更深。

A personal memory layer for your AI tools. Intelligent authoring (gap detection, self-enriching graph, session-synthesis loop) + measurably good retrieval (94.9% blind Cov@10, sub-ms latency). Domain-agnostic framework — works for any context you want to track. MCP-native, Claude-Code-first, works with Cursor/Continue/Cline/OpenAI/Gemini.

Enterprise RAG platform for secure document intelligence using hybrid retrieval, reranking, Gemini LLM, RBAC, HITL, citations, and web search.

On-Premise RAG (продвинутая версия): vLLM/Ollama + bge-m3 + reranker + Qdrant + hybrid (BM25+RRF) + PDR + Basic Auth + Prometheus/Grafana + локальная оценка. Запускается на Ollama без GPU. Тех-задача Petrel AI (Astana Hub).

Verbatim recall across Claude Code sessions, plus a gate against oversized skills. Pure Node, no dependencies, nothing leaves your machine.

基于 DeepSeek Harness 的智能客服工单处理 Agent：自动分类、优先级评估、知识库检索（BM25）与建议答复，结论写回 SQLite 并可全程审计

A retrieval-based medical Q&A system that gives accurate answers to diabetes-related questions by searching a curated medical knowledge base instead of generating text from scratch. Built with a hybrid FAISS + BM25 retrieval pipeline, cross-encoder reranking, and multi-layer safety gating and served through a Streamlit web app.

A Retrieval-Augmented Generation (RAG) chatbot for answering LPDP scholarship questions using Hybrid Retrieval, Cross-Encoder Reranking, and Llama 3.3.

Reproducibility materials for: A Multi-Source RAG Framework for Intelligent Agent Orchestration in Enterprise CRM Systems (IEEE Access 2026)

Document-Driven Aviation RAG Assistant built using FastAPI, FAISS, BM25, Gemini API, and Streamlit. Provides grounded answers from aviation manuals with citations, hallucination prevention, and hybrid retrieval.

Hands-on implementation of Retrieval-Augmented Generation (RAG) using vector databases, embeddings, semantic search, and Anthropic Claude.

End-to-end RAG system built from scratch with recursive chunking, persistent indexing, vector + BM25 hybrid retrieval, evaluation, FastAPI, Docker, and AWS deployment testing.

Enterprise-grade agentic RAG chatbot — Python 3.14 / FastAPI / LangGraph orchestration, PostgreSQL + pgvector hybrid (dense + BM25) retrieval with cross-encoder reranking, Keycloak OIDC auth, MinIO object storage, arq ingestion workers, DeepEval scoring, and a pixel-perfect React 19 + Vite + TypeScript UI.

Apex-relevance subgraph retrieval for AI agents — feed your LLM the most relevant slice of a knowledge graph, within a token budget.