LegalEagleEyeAI 🦅

LegalEagleEyeAI is an agentic AI web application designed to help users understand legal documents and quickly review any risks or obligations before signing. Whether it's a car rental agreement, social media terms, contest rules, or any fine print, LegalEagleEyeAI empowers everyone to make informed decisions with accessible, AI-powered document analysis.

Table of Contents

• Features
• Architecture
• Demo / Video Presentation
• Software Stack
• Environment Variables
• Setup & Local Development
• Usage Guide
• What's Next?
• Attribution
• Contact

---

Features

• Document Upload & Analysis
  Upload PDF, image, or text files. The AI analyzes the document and extracts key clauses, risks, and obligations using advanced LLMs.

• Risk Detection with Severity
  The system highlights and categorizes risk clauses as High Risk, Moderate Risk, or Informational, making it easy to spot critical issues.

• Human-in-the-Loop Review
  Users can review the extracted risks, accept the AI’s analysis, or regenerate a new analysis for more confidence.

• Text-to-Speech (TTS) Accessibility
  The AI can read the summary and risks out loud, using different voices for each severity level. Supports multiple languages for accessibility.

• Translation
  Instantly translate the summary and risk factors into several languages, including Spanish, French, Italian, German, Portuguese, Tamil, and Chinese (Simplified).

• Interactive Q&A Chatbot
  Ask questions about any clause or the full document context. The AI provides clear, context-aware answers.

Architecture

.

Describe the flow:

• Frontend (React): Handles file upload, UI, and user interactions.
• Backend (Flask): Manages agent orchestration, document processing, risk extraction, translation, and TTS.
• Azure Services: OpenAI for LLM, Computer Vision for OCR, Translator, and Speech for TTS.

---

Demo / Video Presentation

https://www.youtube.com/watch?v=8bJk5qFFXPI

---

Software Stack

• Frontend: React, JavaScript, SCSS, Axios
• Backend: Python, Flask, Flask-CORS
• AI & Cloud: Azure OpenAI Service, Azure Computer Vision, Azure Translator, Azure Speech
• PDF/Image Processing: PyMuPDF, pdfminer, requests
• Environment Management: python-dotenv

---

Environment Variables

Before running the project, you must set up several environment variables for Azure and OpenAI services.

Create a .env file in the server/ directory with the following keys:

---

Azure Computer Vision (OCR) AZURE_CV_ENDPOINT=<your_azure_cv_endpoint> AZURE_CV_KEY=<your_azure_cv_key>

Azure OpenAI AZURE_OPENAI_KEY=<your_azure_openai_key> AZURE_OPENAI_ENDPOINT=<your_azure_openai_endpoint> AZURE_OPENAI_DEPLOYMENT=<your_azure_openai_deployment_name>

OpenAI Fallback (optional) OPENAI_API_KEY=<your_openai_api_key>

Azure Speech (Text-to-Speech) SPEECH_KEY=<your_azure_speech_key> SPEECH_REGION=<your_azure_speech_region>

Azure Translator AZURE_TRANSLATOR_KEY=<your_azure_translator_key> AZURE_TRANSLATOR_ENDPOINT=<your_azure_translator_endpoint> AZURE_TRANSLATOR_REGION=<your_azure_translator_region>

---

Setup & Local Development

Prerequisites:

• Node.js >= v22.14
• Python >= 3.12
• Azure resources (OpenAI, Computer Vision, Translator, Speech)

1. Clone the repository

git clone (https://github.com/selvicim45/LegalEagleEyeAI) cd LegalEagleEyeAI

2. Install and Run the Frontend

cd client npm install npm start

• The frontend will run on http://localhost:3000 by default.

3. Install and Run the Backend

cd ../server python3 -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate pip install -r requirements.txt python legal_eagleeye_ai.py

• The backend runs on http://localhost:5001.

4. Environment Variables

• Copy the .env.example to .env in the server/ directory and fill in your Azure/OpenAI credentials.

5. (Optional) PDF/Image Processing

• For best results, ensure your Azure Computer Vision and Translator services are correctly set up for OCR and translation features.

---

Usage Guide

1. Upload a Document:

   • Click "Choose File" and select a PDF, image, or text file containing legal content.

2. Analyze:

   • Click "Upload & Analyze". The AI will extract and summarize key risks and obligations.

3. Review Risks:

   • Review the summary and risk factors. Accept or regenerate the analysis as needed.

4. Accessibility Tools:

   • Use "Read Aloud" to hear the summary and risks, or translate them into your preferred language.

5. Ask Questions:

   • Use the chatbot to ask about any clause or the full document.

---

What's Next?

• User Authentication & Profiles:
  Support for user accounts and saving analysis history.

• Document History & Export:
  Save and export analysis reports as PDF or CSV.

• Customizable Risk Detection:
  Allow users to define custom risk categories or focus areas.

• Multi-Agent Collaboration:
  Enable legal teams to collaborate on document review.

---

Attribution

• Icons:

  • Lucide, Flaticon, Freepik

• Images:

  • Freepik, Unsplash

• Libraries:

  • React, Flask, PyMuPDF, pdfminer, OpenAI, Azure SDKs

---

Contact

Have questions, feedback, or want to contribute?
Contact the development team:

• Arulselvi Amirrthalingam (Selvi) -https://www.linkedin.com/in/arulselvi-amirrthalingam-8450b0193/-

---

LegalEagleEyeAI is designed as a personal legal assistant for everyone. Always consult a qualified professional for critical legal decisions.