Which vector databases does this skill support?

It includes implementation patterns and code templates for Pinecone, Qdrant, and PostgreSQL using the pgvector extension.

Does it support hybrid search?

Yes, the skill provides patterns for combining dense vector search with keyword-based or sparse vector search to improve result relevance.

Is it optimized for large-scale datasets?

Yes, it includes guidance on indexing strategies like HNSW and IVF+PQ that are specifically designed for scaling to millions of vectors with low latency.

Can I use this for RAG applications?

Yes, this skill is specifically designed to help developers implement the retrieval layer of RAG systems, including semantic search and secondary reranking.

What distance metrics are covered?

It provides formulas and use cases for Cosine similarity, Euclidean distance (L2), Dot Product, and Manhattan distance (L1).

Similarity Search Patterns

Name: Similarity Search Patterns
Author: GaitanS

byGaitanS

0•

数据科学与机器学习

Implements high-performance similarity search and vector database patterns for AI-driven applications.

This skill provides standardized implementation patterns for integrating vector databases and similarity search into LLM applications. It offers ready-to-use templates for industry-standard providers like Pinecone, Qdrant, and pgvector, covering essential concepts from distance metrics and indexing strategies to advanced techniques like hybrid search and cross-encoder reranking. It is an essential toolkit for developers building RAG (Retrieval-Augmented Generation) systems, recommendation engines, or semantic search interfaces who need to optimize for both latency and retrieval accuracy in production environments.

主要功能

01Production-ready templates for Pinecone, Qdrant, and pgvector

020 GitHub stars

03Advanced retrieval patterns including hybrid search and cross-encoder reranking

04Best practices for efficient batch upserting and metadata filtering

05Comprehensive guidance on distance metrics including Cosine, L2, and Dot Product

06Detailed indexing strategy comparisons between Flat, HNSW, and IVF

使用场景

01Developing semantic search systems and recommendation engines

02Optimizing vector search performance and recall for large-scale datasets

03Building robust Retrieval-Augmented Generation (RAG) pipelines

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gaitans/ai-mancare similarity-search-patterns

For use in Claude.ai and ChatGPT

Download Skill