Tag
LangChain Community Spotlight featuring Kacper Klusek of Soba Labs on building enterprise RAG by fixing retrieval first: contextual chunking, BM25 hybrid search, and a reranker cut failed retrievals by 67% in Anthropic's evaluation, with LangSmith evals catching document drift.
A developer shares a production-grade RAG assistant built with Python/FastAPI featuring hybrid retrieval (pgvector + Postgres FTS fused via RRF), parent-child chunking, cross-encoder reranking, and Promptfoo-based evals, and is seeking code review and architecture feedback.
一个求职者分享了面试 AI Agent 实施团队时被 RAG 检索细节问题难住的经历，并总结出生产级 RAG 需结合向量检索与 BM25 关键词搜索的混合检索，再用 Cross-encoder Reranker 进行二次精排，以解决语义相近但实体错误的检索问题。
Qdrant publishes a case study on Lucy, which built a source-of-truth retrieval layer over 2.8M+ SEC and Korea DART filings using hybrid lanes and in-query payload filters, achieving 94.7% Recall@10 on FinanceBench and +0.160 answer accuracy with the answer model held fixed.
A technical breakdown of how Papers with Code built a state-of-the-art hybrid search engine combining keyword and semantic search using PostgreSQL, pgvector, and Qwen3 embeddings, powered by Hugging Face infrastructure.
This blog post explains how Hugging Face's Inference Endpoints, Jobs, and Buckets are used to power a hybrid search system for Papers with Code, combining keyword and vector search to improve AI research accessibility.
This paper proposes a hybrid nested search framework that decouples structural sketching (by an LLM) from numeric parameter optimization (by traditional solvers like CMA-ES) in LLM-driven evolutionary optimization, and validates it across meta-optimization, code-based policies, and Bayesian inference tasks.
Agentlas is a free, open-source agent memory system that uses a local SQLite database combined with lightweight embedding and keyword search to achieve 96.4% on LongMemEval without calling an LLM, outperforming alternatives like mem0 and supermemory.
LlamaParse introduces new document retrieval endpoints including hybrid search, file grep, file find, and file read, aiming to improve agentic retrieval quality over unstructured documents.
Cito is a hybrid academic search tool over 236 million papers, designed for AI agents.
Alibaba released Zvec, a fully open-source vector database (Apache 2.0) that can be installed via pip and supports dense, sparse, and hybrid search for RAG applications, processing billions of vectors in milliseconds.
Technical deep dive into Lume, a Rust hybrid search engine that combines BM25, dense vectors, and entity graphs for auditable, local-first retrieval.
A developer shares that the real problem in a RAG app was the retrieval step failing on version numbers and codes, fixed by hybrid search (vector + BM25 + reciprocal rank fusion), not the model.
GBrain is a tool that enhances AI agents with synthesis, gap analysis, a self-wiring typed knowledge graph, hybrid search, and a nightly dream cycle, all built on top of existing .md files in Obsidian.
GBrain is a state-of-the-art retrieval tool for AI agents, released under MIT license, featuring hybrid search, self-wiring knowledge graphs, and temporal question answering, built by Y Combinator's CEO for his own agents.
A developer shares the failure modes encountered while debugging a RAG system, including issues with chunking, stale indices, and hybrid search, along with practical fixes like sliding window chunking and contextual retrieval.
The author argues that most agent RAG failures are due to retrieval problems—specifically chunking errors, lack of freshness signals, and reliance on pure vector search—rather than the LLM, and recommends structural chunking, decay-based ranking, and hybrid BM25+vector search.
Jerry Liu asks about the current state-of-the-art for file search and retrieval, listing options from grep to hybrid search over a database.
This paper describes a system for SemEval-2026 Task 8 that uses a three-stage pipeline involving query rewriting with a fine-tuned Qwen model, hybrid retrieval, and cross-encoder reranking to improve multi-turn retrieval performance.
The article argues against overusing vector search, highlighting BM25's effectiveness for exact keyword matching and its role in hybrid search systems.