Shahadffall/Watheeq-SaudiLegalAssistant

Watheeq: Saudi Legal Assistant ⚖️🇸🇦

Watheeq is an Arabic-language legal chatbot designed to simplify access to Saudi legal documents. It utilizes a Retrieval-Augmented Generation (RAG) pipeline to provide accurate and explainable answers to users' legal questions.

Project Overview

Legal documents in Saudi Arabia can be complex and hard to navigate. Watheeq makes it easier by:

• —Answering legal questions in Arabic
• —Retrieving information from official Saudi laws (scraped from هيئة الخبراء)
• —Using plain language to explain laws and procedures
• —Supporting the Saudi Vision 2030 goals for digital transformation

How It Works

This notebook implements a full RAG pipeline:

1. 1.Data Loading Loads .txt files containing Saudi laws (e.g., labor laws, legal procedures).

1. 1.Chunking Splits long documents into 500-character chunks with 50-character overlap for better context during retrieval.

1. 1.Embeddings Uses CohereEmbeddings to create vector representations of legal text.

1. 1.Vector Store (ChromaDB) Stores embedded documents in a persistent ChromaDB store for efficient similarity search.

1. 1.LLM with Groq API Queries are answered using Llama3 via Groq for fast and high-quality response generation.

1. 1.Chain Construction Combines document retriever + prompt template + LLM into a LangChain RetrievalQA pipeline.

Example Use Case

Ask a question like:

"ما مضمون المادة السابعة من نظام العمل؟"

The system will:

• —Search for relevant text chunks
• —Send the context to the LLM
• —Return a grounded, natural-language answer

📁 Folder Structure

Watheeq_Final.ipynb     # Full notebook with all code
/laws                   # Folder containing .txt legal files
/chroma_laws_db         # Vector store directory (generated)