Transforming Invoice Management with Artificial Intelligence
📺 Demo Video • 📋 Problem Statement • 🚀 Quick Start • 📖 Documentation
- About
- Problem Statement
- Our Solution
- Key Features
- Tech Stack
- Getting Started
- Demo
- Team
- Acknowledgments
LUMEN is an intelligent invoice management system built for Hack-a-Sol Hackathon at IIIT Naya Raipur by Team dUnder Pressure. The project addresses the critical challenge of manual invoice processing in modern businesses by leveraging cutting-edge AI technologies.
- Event: Hack-a-Sol Hackathon
- Institution: Indian Institute of Information Technology, Naya Raipur
- Team: dUnder Pressure
- Challenge: View Problem Statement
See PROBLEM_STATEMENT.md for the complete problem statement.
Manual invoice processing is a time-consuming, error-prone task that costs businesses significant resources. Finance teams spend 8+ hours weekly on:
- Manual data entry from invoices
- Cross-verification and validation
- Payment tracking and reminders
- Spending analysis and reporting
- Fraud detection and compliance
The Challenge: Build an AI-powered solution that automates invoice processing, provides intelligent analytics, and enables natural language querying of financial data.
LUMEN transforms invoice management through three core pillars:
- 95%+ accuracy in invoice data extraction
- Support for images, PDFs, and multi-page documents
- Automatic validation and error detection
- Real-time processing with instant feedback
- Pattern Detection: Identifies spending trends and seasonal patterns
- Anomaly Detection: Flags suspicious transactions and potential fraud
- Forecasting: Predicts future spending with ML models
- Risk Assessment: Evaluates vendor and payment risks
- Smart Insights: AI-generated recommendations for cost savings
- Ask questions in plain English
- Hybrid SQL + RAG system for comprehensive answers
- Context-aware responses with data citations
- Query classification for optimal routing
- Drag & Drop Upload: Simple file upload interface
- Batch Processing: Handle multiple invoices simultaneously
- Auto-validation: Intelligent field validation
- Database Storage: Secure PostgreSQL storage
- Export Options: Excel, CSV, PDF reports
- Real-time KPIs: Annual spending, monthly averages, overdue payments
- Spending Trends: Interactive graphs with forecasts
- Category Breakdown: Pie charts and detailed analysis
- Payment Calendar: Visual payment schedule
- Alert System: Automated reminders and notifications
- Conversational Interface: Natural language queries
- Multi-agent System: SQL Agent + RAG System + Analytics Orchestrator
- Query Examples:
- "Show me all invoices from last month"
- "Which vendor has the highest spending?"
- "Are there any unusual transactions?"
- "What's my projected spending for next quarter?"
- Dark Mode: Elegant dark theme throughout
- Responsive Design: Works on desktop, tablet, and mobile
- Animations: Smooth Framer Motion animations
- Accessibility: WCAG compliant components
- Intuitive Navigation: Easy-to-use sidebar and breadcrumbs
- Framework: Next.js 14.2.5 (React 18)
- Language: TypeScript
- Styling: Tailwind CSS
- UI Components: Radix UI + Shadcn/ui
- Animations: Framer Motion
- Charts: Recharts
- Icons: Lucide React, Tabler Icons
- HTTP Client: Axios
- Framework: Flask 3.0.0
- Language: Python 3.11+
- Database: SQLite (dev) / PostgreSQL (production)
- ORM: SQLAlchemy with Alembic migrations
- AI/ML:
- OpenRouter API (LLM integration)
- Google Cloud Vision API (OCR)
- Sentence Transformers (embeddings)
- ChromaDB (vector database)
- Scikit-learn (ML models)
- Query Classifier: Routes queries to appropriate systems
- SQL Agent: Generates and executes SQL queries
- RAG System: Retrieval-Augmented Generation for document Q&A
- Pattern Detection Agent: Identifies spending patterns
- Anomaly Detection Agent: Fraud and anomaly detection
- Forecasting Agent: Time-series forecasting
- Risk Assessment Engine: Vendor and payment risk analysis
- Version Control: Git & GitHub
- API Testing: Postman
- Environment: Virtual environments (venv)
- CORS: Flask-CORS for cross-origin requests
- Node.js 18+ and npm
- Python 3.11+
- Git
# Clone the repository
git clone https://github.com/CoderUzumaki/Lumen.git
cd Lumen📚 Complete Documentation:
- Quick Start Guide - Get running in 5 minutes
- Detailed Setup Guide - Comprehensive installation instructions
- API Documentation - Complete API reference
- Architecture Guide - Technical deep dive
- Deployment Guide - Production deployment
- Contributing Guide - How to contribute
- Documentation Index - Navigate all docs
Note: Replace
YOUR_VIDEO_IDwith your actual YouTube video ID after uploading the demo.
Demo Highlights:
- 📄 Invoice upload and OCR extraction
- 📊 AI analytics dashboard walkthrough
- 💬 Natural language query examples
- 🚨 Anomaly detection in action
- 📈 Forecasting and insights
- 🎯 End-to-end workflow demonstration
Team dUnder Pressure
Built with dedication by our amazing team for Hack-a-Sol @ IIIT Naya Raipur:
| Member | Role | Responsibilities | GitHub |
|---|---|---|---|
| [Your Name] | Full Stack Lead | Frontend architecture, UI/UX, Landing page | @CoderUzumaki |
| [Team Member 2] | AI/ML Engineer | AI agents, ML models, Analytics | @username2 |
| [Team Member 3] | Backend Developer | API development, Database, OCR integration | @username3 |
| [Team Member 4] | Designer & Frontend | Component design, Animations, Testing | @username4 |
Note: Replace team member names and GitHub usernames with your actual team information.
- Hack-a-Sol Organizing Committee at IIIT Naya Raipur
- Mentors and Judges for their guidance and feedback
- Open Source Community for the amazing tools and libraries
- Next.js - React framework
- Flask - Python web framework
- OpenRouter - LLM API gateway
- Google Cloud Vision - OCR API
- ChromaDB - Vector database
- Vercel - Frontend hosting
- Render - Backend hosting
This project was created for the Hack-a-Sol Hackathon at IIIT Naya Raipur.
For questions or feedback about this project:
- Email: team@dunder-pressure.dev
- GitHub Issues: Report a bug
- Project Maintainer: @CoderUzumaki
Built with ❤️ by Team dUnder Pressure for Hack-a-Sol @ IIIT Naya Raipur
⭐ Star us on GitHub if you find this project helpful!