Eigen - AI-Powered Real Estate Investment and Finance Platform

A comprehensive full-stack application that combines AI-powered document analysis with real estate investment screening. Built with FastAPI backend and Next.js frontend, featuring advanced LangChain integration for intelligent property data extraction and analysis.

✨ Key Features

🤖 AI-Powered Analysis

  • Intelligent Document Processing: Automatically extracts property data from various document formats (PDF, Excel, CSV, PowerPoint, etc.)
  • Data-Driven Screening: AI agent with emphasis on facts and quantitative analysis
  • Memory-Based Search: Vector database for intelligent document retrieval and context
  • Multi-Format Support: Handles PDFs, Excel files, PowerPoint presentations, Word documents, and more

📊 Real Estate Investment Tools

  • Property Data Extraction: Structured extraction of financial metrics, property details, and market data
  • Investment Screening: Comprehensive analysis of real estate opportunities
  • Risk Assessment: Data-driven risk evaluation and mitigation strategies

🔧 Technical Features

  • Modern Tech Stack: FastAPI + Next.js 15 with TypeScript
  • AI Integration: LangChain with Google Gemini 2.5 Pro
  • Vector Database: ChromaDB for document memory and search
  • Real-time Chat: Interactive AI agent with conversation context
  • File Processing: OCR, document parsing, and data extraction
  • Authentication: JWT-based security system

🤖 AI Agent Commands

The AI agent supports special commands for enhanced functionality:

  • @screener - Run comprehensive screening on all documents
  • @memory [query] - Search memory for specific information
  • @stats - Show memory statistics and document counts
  • @help - Display available commands

Example Usage

User: @screener
AI: 🔍 **MEMORY SCREENING RESULTS** [Detailed analysis of all documents]

User: @memory market trends
AI: 🔍 **Memory Search Results** [Relevant market trend information]

User: What's my portfolio performance?
AI: [Data-driven analysis based on uploaded documents]

📊 Supported Document Types

The system can process and extract data from:

  • PDFs - Property reports, financial statements
  • Excel/CSV - Financial data, market analysis
  • PowerPoint - Investment presentations
  • Word Documents - Property descriptions, legal docs
  • RTF/ODT - Alternative document formats
  • Text Files - Raw data and notes

📈 Performance Features

  • Optimized Document Processing: Efficient parsing with content truncation for large files
  • Caching: Redis-based caching for improved response times
  • Vector Search: Fast semantic search using ChromaDB
  • Async Processing: Non-blocking operations for better user experience
  • Memory Management: Intelligent document memory with cleanup

🎯 Roadmap

  • [ ] Real-time collaboration features
  • [ ] Advanced portfolio analytics
  • [ ] Integration with real estate APIs
  • [ ] Mobile application
  • [ ] Advanced AI model fine-tuning
  • [ ] Multi-tenant support

Built with ❤️ for VTHacks13

Built With

Share this project:

Updates