rag-engineer

The RAG Engineer skill specializes in architecting Retrieval-Augmented Generation (RAG) systems, bridging the gap between raw documents and LLM understanding. It focuses on optimizing retrieval quality, which directly impacts generation quality, ensuring helpful and accurate LLM responses.

This skill encompasses expertise in document chunking, embedding models, vector databases, and retrieval pipelines. It implements strategies like semantic chunking, hierarchical retrieval, and hybrid search to improve precision and recall. The skill also addresses common anti-patterns and sharp edges in RAG system design, offering solutions for issues like fixed-size chunking and embedding evaluation.

RAG Engineer works to create semantic search, document retrieval, and vector search systems for diverse LLM applications.

What It Does

Builds and optimizes Retrieval-Augmented Generation (RAG) systems for LLMs. This includes document preprocessing, embedding generation, vector database management, and retrieval pipeline design.

When To Use

When building RAG systems, implementing vector search, creating embeddings, implementing semantic search, or optimizing document retrieval for LLM applications.

Installation

Copy SKILL.md to your skills directory