Legal_AI

akhil-katta1/Legal_AI

3.2

If you are the rightful owner of Legal_AI and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The Legal AI System is designed to assist attorneys in generating demand letters and analyzing cases efficiently.

Tools
5
Resources
0
Prompts
0

Legal AI System

Overview

This project is a Legal AI Assistant designed to help attorneys streamline demand letter generation and case analysis. It integrates:

  • A PostgreSQL case management system (MCP server)
  • A Retrieval-Augmented Generation (RAG) pipeline for document search and citation
  • LLM-powered demand letter generation with inline citations
  • A frontend UI for attorney interaction

Features

  • Retrieve structured case details via MCP endpoints
  • Ingest, chunk, and embed legal documents into ChromaDB
  • Perform semantic search with page-level citations
  • Extract dates, financial amounts, and named entities
  • Generate demand letters with citations and validation
  • Save outputs as .txt and .docx files

System Architecture

flowchart TD
    A[Attorney Request] --> B["UI: Streamlit / FastAPI"]
    B --> C["MCP Server: FastAPI + PostgreSQL"]
    C --> D["Database: Cases, Parties, Events"]
    C --> E["Case Documents: PDFs"]
    E --> F["RAG Pipeline: Embeddings + ChromaDB"]
    F --> G["LLM Layer: GPT-4 / LegalBERT"]
    G --> H["Demand Letter Generator"]
    H --> I["Output: Word / TXT / UI Preview"]

Setup Instructions

1. Clone Repository

git clone https://github.com/your-username/legal-ai-system.git
cd legal-ai-system

2. Create Virtual Environment

python -m venv venv
source venv/bin/activate   # Linux/Mac
venv\Scripts\activate      # Windows

3. Install Requirements

pip install -r requirements.txt

4. Configure Environment

Create a .env file:

OPENAI_API_KEY=your-openai-key
DB_NAME=legal_case_management
DB_USER=postgres
DB_PASSWORD=yourpassword
DB_HOST=127.0.0.1
DB_PORT=5432

5. Setup Database

Run schema script in PostgreSQL (see provided SQL file). Verify with:

\c legal_case_management;
SELECT * FROM cases;

6. Run MCP Server

uvicorn mcp.server:app --reload

7. Ingest Documents

python rag/legal_ai_system.py

8. Generate Demand Letter

python generator/demand_letter.py --request "Generate a demand letter for Case #2024-PI-001 focusing on medical expenses and lost wages"

Output:

  • sample_output/demand_letter_2024-PI-001.txt
  • sample_output/demand_letter_2024-PI-001.docx

9. Run Frontend UI

streamlit run app/ui.py

MCP Endpoints

  1. get_case_details(case_id) → Case info + parties
  2. get_case_documents(case_id, category=None) → Document metadata
  3. get_case_timeline(case_id, event_type=None) → Chronological events
  4. get_financial_summary(case_id) → Damages summary
  5. search_similar_cases(case_type, keywords) → Case precedents
  6. get_party_details(case_id, party_type) → Specific party info

Deliverables

  • ✅ Codebase with structured repo
  • ✅ Working MCP server with DB integration
  • ✅ RAG pipeline with citations
  • ✅ Demand letter generator with QA validation
  • ✅ Sample outputs in /sample_output/
  • ✅ README documentation

Example Attorney Requests

  • "Generate a demand letter for Case #2024-PI-001 focusing on medical expenses and lost wages."
  • "Generate a demand letter for Case #2024-PI-001. Include all medical expenses, lost wages, and pain and suffering damages. Reference Dr. Jones and cite the police report."