📘 VerdictIQ – Legal Case Analysis System

VerdictIQ is a full-stack legal tech platform designed to make legal analysis faster, fairer, and more accessible, particularly for the Indian justice system. It leverages machine learning, NLP, and XAI techniques (like SHAP/LIME) to classify legal cases, predict case outcomes, summarize legal documents, and suggest legal strategies via an interactive web interface.

---

🚀 Key Features

• ⚖️ ML-based Legal Case Classification (e.g., civil, criminal, family, corporate)
• 📊 Outcome Prediction using court, city, year, and case description with visual probabilities
• 📚 Legal Strategy Advisor based on past precedent analysis
• 📄 Legal Document Summarization & Extraction via Hugging Face NLP models
• 🔍 Precedent Search & Comparison with filters for Indian jurisdictions
• 🌐 Multilingual Document Support for regional legal texts
• 🛡️ Fairness Monitoring (bias detection & explainability with SHAP/LIME)
• 🎯 For Students, Lawyers, Judges, and Citizens to democratize legal access

---

🛠️ Prerequisites

Ensure the following are installed:

• Node.js (v14+)
• Python (v3.8+)
• pip
• Git

Recommended Specs:
Minimum 4GB RAM, and open ports: 3000, 8001, 8002.

---

🔧 Step-by-Step Setup Guide

1. Clone the Repository

git clone <repository-url>
cd <repository-folder>

---

2. Setup Frontend (React)

cd verdictiq
npm install

---

3. Setup Backend ML Services

⚠️ Repeat the below steps separately for both myModel-classification and myModel-Prediction

A. For Classification Model (Port 8002)

cd ../myModel-classification
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt

B. For Outcome Prediction Model (Port 8001)

cd ../myModel-Prediction
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt

---

▶️ Running the Project (Development Mode)

1. Start Backend Services

Classification API (Port 8002)

cd myModel-classification
source venv/bin/activate
python app.py

Prediction API (Port 8001)

cd ../myModel-Prediction
source venv/bin/activate
python app.py

Verify Servers:

• http://localhost:8001 → Outcome Prediction
• http://localhost:8002 → Case Classification

---

2. Start Frontend (Port 3000)

cd ../verdictiq
npm start

Visit the app: http://localhost:3000

---

📚 Backend Dependencies

Installed via pip install -r requirements.txt:

• fastapi
• uvicorn
• scikit-learn
• pandas
• numpy
• joblib
• transformers
• shap, lime
• others as needed

---

🚀 Build for Production

1. Build Frontend

cd verdictiq
npm run build

2. Deploy Backend ML Services

Recommended:

• Use Gunicorn or Uvicorn with ASGI
• Set up Nginx or Apache as reverse proxy
• Add SSL Certificates (Let's Encrypt)
• Define secure .env variables for configs
• Use PM2 or Docker for process management

---

🛡️ Troubleshooting

Issue	Solution
Port already in use	Kill the process or change the port
Model not loading	Ensure model.pkl and vectorizer.pkl exist
API not connecting	Ensure the backend servers are running
Frontend not displaying	Check browser console and verify backend connectivity

---

👥 Target Audience

• Law Students & Researchers
• Lawyers & Legal Analysts
• Judges & Court Assistants
• Citizens seeking fair legal insights

---

🧪 Technologies Used

• Frontend: React.js, Tailwind CSS, Vite
• Backend: Python, FastAPI, scikit-learn, Hugging Face Transformers
• ML/NLP: BERT, SHAP, LIME, TF-IDF, Logistic Regression
• Infra: MongoDB, GitHub, REST APIs

---

📦 Planned Enhancements

• Public Deployment (Vercel, Render, or Railway)
• Expanded Indian legal datasets (Indian Kanoon, Supreme Court)
• Paid tier for legal professionals
• In-app Legal Chatbot (RAG + PDF memory)
• Better Precedent Filtering and PDF Export Tools

---

🤝 Contributing

Contributions are welcome! Please open an issue or submit a pull request for:

• Bug fixes
• Feature requests
• Model improvements
• Language support additions

---

📄 License

This project is open-source under the MIT License.