frutik/awesome-search

Awesome Search - this is all about the (e-commerce, but not only) search and its awesomeness

What it solves

This project provides a comprehensive, curated knowledge base for developers and engineers building search applications, particularly in e-commerce. It organizes 15+ years of industry experience into a structured map of concepts, tools, and research to help navigate the complex landscape of modern information retrieval.

How it works

It functions as a "map of content" or a digital library. The project categorizes technical notes into specific themes—such as lexical retrieval, dense embeddings, vector indexing, and generative search—allowing users to explore the relationships between different search technologies, academic papers, and practical implementation strategies.

Who it’s for

This resource is designed for search engineers, data scientists, and developers who need a centralized reference for both foundational search principles and cutting-edge developments like RAG (Retrieval-Augmented Generation) and agentic search.

Highlights

  • Extensive coverage of Lexical and Sparse Retrieval methods like BM25 and SPLADE.
  • Deep dive into Dense Retrieval, including embedding techniques and vector similarity metrics.
  • Detailed sections on ANN (Approximate Nearest Neighbor) indexing and quantization strategies.
  • Comprehensive guides on Reranking, Learning to Rank (LTR), and Hybrid Search.
  • Specialized modules for Query Understanding, Evaluation Metrics, and Agentic/Generative search architectures.

Related

  • Project
    CHIANGEL/Awesome-LLM-for-RecSys

    A curated collection of academic papers and resources exploring the integration of Large Language Models into recommender systems, categorized by their role in the recommendation pipeline.

  • Project
    firelex/jeff

    Jeff is a 0.8 B “System 1” model that makes ultra‑fast, calibrated classification decisions. It uses lightweight LoRA adapters for specific domains (guard, triage, spam, etc.) and can act as a fast‑path filter in front of a large LLM (e.g., Qwen 3.8‑27B), handling most queries in ~30 ms and delegating only uncertain cases. Weights and adapters are open on Hugging Face, GGUF files enable llama.cpp use, and a simple HTTP/Python API (`jeff‑serve`) lets you plug Jeff into any service.

  • Project
    finic-ai/rag-stack

    RAGstack is a legacy, self-hosted platform for deploying private Retrieval Augmented Generation (RAG) applications that allow users to chat with their own documents using open-source LLMs.

  • Project
    shibing624/similarity

    A Java toolkit for computing similarity scores between words, phrases, sentences, and paragraphs, featuring sentiment analysis and Word2vec-based synonym discovery.