RAG Implementation

Master Retrieval-Augmented Generation (RAG) to build LLM applications that provide accurate, grounded responses using external knowledge sources.

When to Use This Skill

• Building Q&A systems over proprietary documents
• Creating chatbots with current, factual information
• Implementing semantic search with natural language queries
• Reducing hallucinations with grounded responses
• Enabling LLMs to access domain-specific knowledge
• Building documentation assistants
• Creating research tools with source citation

Core Components

1. Vector Databases

Purpose: Store and retrieve document embeddings efficiently

Options:

• Pinecone: Managed, scalable, serverless
• Weaviate: Open-source, hybrid search, GraphQL
• Milvus: High performance, on-premise
• Chroma: Lightweight, easy to use, local development
• Qdrant: Fast, filtered search, Rust-based
• pgvector: PostgreSQL extension, SQL integration

2. Embeddings

Purpose: Convert text to numerical vectors for similarity search

Models (2026):

3. Retrieval Strategies

Approaches:

• Dense Retrieval: Semantic similarity via embeddings
• Sparse Retrieval: Keyword matching (BM25, TF-IDF)
• Hybrid Search: Combine dense + sparse with weighted fusion
• Multi-Query: Generate multiple query variations
• HyDE: Generate hypothetical documents for better retrieval

4. Reranking

Purpose: Improve retrieval quality by reordering results

Methods:

• Cross-Encoders: BERT-based reranking (ms-marco-MiniLM)
• Cohere Rerank: API-based reranking
• Maximal Marginal Relevance (MMR): Diversity + relevance
• LLM-based: Use LLM to score relevance

Quick Start with LangGraph

from langgraph.graph import StateGraph, START, END
from langchain_anthropic import ChatAnthropic
from langchain_voyageai import

Detailed patterns and worked examples

Detailed pattern documentation lives in references/details.md. Read that file when the navigation tier above is insufficient.