Policy Brain - LLM-Powered Intelligent Query-Retrieval System

A distributed application for intelligent document analysis using AI-powered insights across Insurance, Legal, HR, and Compliance domains.

Project Structure

This is a monorepo containing both frontend and backend applications:

policy-brain/
├── frontend/          # Next.js React application
│   ├── app/          # Next.js app router pages
│   ├── components/   # Reusable React components
│   ├── lib/          # Client-side utilities and API calls
│   ├── public/       # Static assets
│   └── package.json  # Frontend dependencies
├── backend/           # Node.js Express API server
│   ├── src/          # TypeScript source code
│   ├── dist/         # Compiled JavaScript
│   ├── uploads/      # File upload directory
│   └── package.json  # Backend dependencies
└── package.json      # Monorepo root with shared scripts

Features

• Multi-Format Document Support: Upload and process PDF, DOCX, and TXT files
• AI-Powered Analysis: Intelligent document analysis with semantic search
• Vector Embeddings: Local sentence-transformer embeddings for document chunks
• LLM Integration: Groq Cloud LLaMA 3 for contextual answer generation
• Vector Database: Pinecone for efficient similarity search
• Domain-Specific Intelligence: Tailored for Insurance, Legal, HR, and Compliance

Prerequisites

• Node.js 18+
• npm or yarn
• Pinecone account and API key
• Groq Cloud API key

Installation

1. Clone and install dependencies:

   git clone <repository-url>
   cd policy-brain
   npm install
   npm run install:all

2. Set up environment variables:

   Backend (.env):

   # Environment Configuration
   NODE_ENV=development
   PORT=5001
   
   # Pinecone Configuration
   PINECONE_API_KEY=your_pinecone_api_key
   PINECONE_INDEX_NAME=policy-brain-index
   PINECONE_ENVIRONMENT=us-east-1
   
   # Groq Configuration
   GROQ_API_KEY=your_groq_api_key
   GROQ_MODEL=llama3-8b-8192
   
   # File Upload Configuration
   MAX_FILE_SIZE=10485760
   UPLOAD_DIR=uploads
   
   # Embedding Configuration
   CHUNK_SIZE=1000
   CHUNK_OVERLAP=200

   Frontend (.env.local):

   NEXT_PUBLIC_BACKEND_URL=http://localhost:5001

Development

Running Both Services Together

npm run dev

This will start both frontend (http://localhost:3000) and backend (http://localhost:5001) concurrently.

Running Services Individually

Frontend only:

npm run dev:frontend
# or
cd frontend && npm run dev

Backend only:

npm run dev:backend
# or
cd backend && npm run dev

API Endpoints

• Health Check: GET /health
• Upload Document: POST /api/upload
• Query Documents: POST /api/query
• List Documents: GET /api/documents

Building for Production

# Build both services
npm run build

# Start production servers
npm run start

Technology Stack

Frontend

• Next.js 15 - React framework with App Router
• React 19 - UI library
• TypeScript - Type safety
• Tailwind CSS - Styling
• Lucide React - Icons

Backend

• Node.js - Runtime
• Express.js - Web framework
• TypeScript - Type safety
• Pinecone - Vector database
• Groq Cloud - LLM API
• Xenova Transformers - Local embeddings

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Run tests and linting
5. Submit a pull request

License

MIT License - see LICENSE file for details.

Backend (.env):

PINECONE_API_KEY=your_pinecone_api_key
GROQ_API_KEY=your_groq_api_key
PINECONE_INDEX_NAME=policy-brain-index
PINECONE_ENVIRONMENT=us-east-1

Frontend (.env.local):

NEXT_PUBLIC_BACKEND_URL=http://localhost:5001

Development

Start both services:

npm run dev

This will start:

• Frontend: http://localhost:3005
• Backend: http://localhost:5001

Start services individually:

Frontend only:

npm run dev:frontend

Backend only:

npm run dev:backend

API Endpoints

Document Upload

POST /api/upload
Content-Type: multipart/form-data

Upload documents for processing and embedding.

Query Documents

POST /api/query
Content-Type: application/json

Query documents with natural language questions.

System Status

GET /api/status

Get system health and statistics.

Building for Production

npm run build

Deployment

Frontend (Vercel)

cd frontend
npm run build
npm run start

Backend (Railway, Heroku, etc.)

cd backend
npm run build
npm run start

Tech Stack

Frontend

• Framework: Next.js 15 with App Router
• Language: TypeScript
• Styling: Tailwind CSS
• Icons: Lucide React

Backend

• Runtime: Node.js with Express.js
• Language: TypeScript
• AI/ML: Groq SDK, @xenova/transformers
• Database: Pinecone Vector Database
• File Processing: pdf-parse, mammoth
• Security: Helmet, CORS

Contributing

1. Install dependencies: npm run install:all
2. Start development: npm run dev
3. Make changes to frontend/ or backend/
4. Test both services work together
5. Submit pull request

License

This project is licensed under the MIT License.