📜 Legal Question Generator

A Streamlit app that reads Indian constitutional law case briefs (PDFs) and generates scenario-based multiple choice questions (MCQs) using an LLM via the Groq API.

This project explores agentic AI—systems that perform multi-step tasks independently—and applies it to legal education.

---

🧠 What It Does

• 📂 Upload a legal case brief (PDF)
• 🤖 Extracts and cleans the text
• 🔍 Sends a structured prompt to a Groq-hosted LLaMA model
• ❓ Generates MCQs focused on Article 21 themes (e.g., life, liberty, privacy)
• ✅ Displays correct answers with legal reasoning
• 📥 Export results in Markdown or plain text

---

🛠 Tech Stack

• Python + Streamlit
• pdfplumber + PyPDF2 for PDF parsing
• Groq API with LLaMA-4
• Simple regex-based output parser

---

🚀 Getting Started

1. Clone the repo:

   git clone https://github.com/yourusername/legal-question-generator.git
   cd legal-question-generator

2. Install dependencies:

   pip install -r requirements.txt

3. Add your Groq API key:
   Set it as an environment variable or add to .streamlit/secrets.toml:

   GROQ_API_KEY = "your_api_key_here"

4. Run the app:

   streamlit run app.py

---

📌 Notes

• Currently optimized for Indian constitutional law content
• Focuses on Article 21 cases like Maneka Gandhi, Puttaswamy, D.K. Basu, etc.
• Not production-ready — built for experimentation and learning

---

🧪 Why This?

To practice building small agentic AI tools that chain together:

• PDF extraction → prompt engineering → model response → structured output → user interaction

---

📄 License

MIT License