skills.rest
enigmaicon-eng
enigmaicon-engCOMMUNITY
Explore all skills by this author →
GitHub

rag-architect

Design production RAG pipelines with chunking, embeddings, and vector store indexing.

1|Updated May 16, 2026
Search Tags:#semantic search#embeddings#rag#vector database#reranking#hybrid retrieval#retrieval evaluation
One-click install
npx skills add https://github.com/enigmaicon-eng/AI-Enterprise-OS --skill rag-architect-enigmaicon-eng
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-architect
Source: https://github.com/enigmaicon-eng/AI-Enterprise-OS/tree/main/examples/claude-skills/skills/rag-architect
Command: npx skills add https://github.com/enigmaicon-eng/AI-Enterprise-OS --skill rag-architect-enigmaicon-eng

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you stop building brittle, low-quality RAG systems by providing a production-oriented blueprint for designing retrieval that actually finds the right context.

Core Features & Use Cases

  • Chunking that respects semantics: choose and implement chunking strategies (size, overlap, and boundaries) to improve retrieval relevance.
  • Embedding + vector store architecture: generate embeddings, design indexing/deduplication, and configure vector databases for reliable similarity search.
  • Hybrid retrieval + reranking: combine dense and sparse retrieval, apply reranking to top results, and evaluate retrieval quality using standard metrics.
  • Evaluation and iteration loop: measure retrieval metrics (precision/recall/MRR/NDCG) and tune before integrating with the LLM.

Quick Start

Ask the AI to generate a production RAG retrieval architecture for my knowledge base, including chunking parameters, embedding and vector DB setup, hybrid search with reranking, and an evaluation plan with target thresholds.

Frequently Asked Questions about rag-architect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I design a production RAG pipeline for enterprise knowledge bases?▼

Build a reliable RAG pipeline by implementing semantic chunking, embedding generation, and vector store indexing. Support enterprise knowledge bases by adding metadata enrichment and multi-tenant domain filtering for precise context retrieval.

What is hybrid retrieval and how does reranking improve RAG search results?▼

Hybrid retrieval combines dense vector similarity search with sparse keyword matching to capture both semantic and exact-match context. Reranking refines the top-k retrieved results to maximize precision before passing context to the LLM.

How do I evaluate retrieval quality using precision, recall, MRR, and NDCG?▼

Evaluate retrieval quality by measuring precision@k and recall@k for exact top-k accuracy, MRR for first relevant result ranking, and NDCG for graded relevance. Tune chunking and embedding parameters against these metrics before integrating the LLM.

What is the best way to chunk documents for semantic search?▼

Chunk documents for semantic search by configuring chunk size, overlap, and structural boundaries to respect semantic context. This prevents splitting related information and improves embedding generation and retrieval relevance.

How do I set up vector database indexing for embeddings?▼

Set up vector database indexing by generating embeddings from chunked documents and configuring indexing and deduplication parameters. This ensures reliable similarity search and fast retrieval for production-grade retrieval-augmented generation pipelines.

When should I not use dense vector search alone for retrieval?▼

Avoid using dense vector search alone when your knowledge base requires exact keyword matching or multi-tenant domain filtering. Implement hybrid retrieval combining dense and sparse retrieval, followed by reranking, to capture exact matches and semantic context.

Related Skills

Explore More Software Engineering Skills→
patricio0312revpatricio0312rev
5

embedding-pipeline-builder

Automate document embedding pipelines for semantic search.

Community
Advanced
marchattonmarchatton
5

evaluate-rag

Evaluate RAG pipelines with Recall@k, Precision@k, MRR, and NDCG@k metrics.

Community
Advanced
JFrangelJFrangel

chunk-scoring

Optimize RAG pipelines with semantic chunking, embedding selection, and vector database configuration.

Community
Advanced
zilliztechzilliztech
3

hybrid-search

Combine BM25 keyword matching with vector semantic similarity in Milvus.

Official
Advanced