Which vector databases are covered by this skill?

The skill includes implementation details for Pinecone, Weaviate, Qdrant, Chroma, and pgvector (PostgreSQL).

What chunking strategies does RAG Architect support?

It provides guidance on fixed-size, sentence-based, paragraph-based, semantic, recursive, and document-aware chunking to ensure optimal context preservation.

How does this skill improve RAG accuracy?

It utilizes advanced techniques such as HyDE (Hypothetical Document Embeddings), multi-query generation, and reranking to increase retrieval precision.

Does it help with RAG evaluation?

Yes, it incorporates frameworks like RAGAS to measure faithfulness, relevance, and context precision in production environments.

RAG Architect

Name: RAG Architect
Author: zhangzhang-111-i

byzhangzhang-111-i

0•

Data Science & ML

Designs, implements, and optimizes production-grade Retrieval-Augmented Generation (RAG) pipelines for AI applications.

The RAG Architect skill provides Claude with specialized knowledge to build sophisticated retrieval systems, covering the end-to-end lifecycle from document chunking and embedding selection to advanced retrieval strategies and performance evaluation. It offers architectural guidance for selecting vector databases, implementing hybrid search, and optimizing context windows to ensure AI responses are accurate, grounded, and cost-effective. Whether you are building an enterprise knowledge base or a simple semantic search engine, this skill provides the patterns needed for high-performance retrieval.

Key Features

01Comparative analysis for vector databases like Pinecone, Weaviate, and pgvector

02Query transformation patterns like HyDE and Multi-Query Generation

03Performance evaluation frameworks using RAGAS and faithfulness metrics

04Sophisticated retrieval techniques including hybrid search and reranking

050 GitHub stars

06Advanced document chunking strategies including semantic and recursive splitting

Use Cases

01Architecting cost-effective vector storage and embedding strategies for large datasets

02Optimizing existing RAG pipelines to improve response accuracy and reduce hallucinations

03Building enterprise-scale knowledge bases with accurate context retrieval

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhangzhang-111-i/claude-skills111 rag-architect

For use in Claude.ai and ChatGPT

Key Features

01Comparative analysis for vector databases like Pinecone, Weaviate, and pgvector

02Query transformation patterns like HyDE and Multi-Query Generation

03Performance evaluation frameworks using RAGAS and faithfulness metrics

04Sophisticated retrieval techniques including hybrid search and reranking

050 GitHub stars

06Advanced document chunking strategies including semantic and recursive splitting

Use Cases

01Architecting cost-effective vector storage and embedding strategies for large datasets

02Optimizing existing RAG pipelines to improve response accuracy and reduce hallucinations

03Building enterprise-scale knowledge bases with accurate context retrieval

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhangzhang-111-i/claude-skills111 rag-architect

For use in Claude.ai and ChatGPT