lex-rag-builder

What problem does it solve?

This Skill addresses the challenges of building effective Retrieval-Augmented Generation (RAG) systems specifically for Brazilian legal documents, ensuring accuracy and relevance in AI-driven legal research and knowledge management.

Core Features & Use Cases

• Specialized Chunking: Implements strategies for parsing legal documents like court decisions and legislation.
• Vector Store Configuration: Provides examples for Qdrant and Supabase pgvector.
• Embedding Model Selection: Recommends and ranks embedding models suitable for Portuguese legal text.
• Query Reformulation: Enhances search recall with legal-specific term expansion.
• RAG Evaluation: Outlines metrics for assessing RAG quality in a legal context.
• Use Case: You need to build a knowledge base for an AI assistant that can answer questions about Brazilian civil law using a collection of court rulings and statutes. This skill will guide you through setting up the data ingestion, vector storage, and retrieval process.

Quick Start

Use the lex-rag-builder skill to set up a Qdrant vector store for Brazilian legal documents.