Skip to content

replit.md

Latest commit

 

History

History
143 lines (114 loc) · 6.68 KB

replit.md

File metadata and controls

143 lines (114 loc) · 6.68 KB

Legal Document Analysis Platform

Overview

This is a full-stack legal document analysis platform that allows users to search, analyze, and visualize relationships between legal documents, names, and entities. The application provides a dashboard for exploring court filings, flight logs, contact lists, and other legal documents while tracking connections between mentioned individuals.

User Preferences

Preferred communication style: Simple, everyday language.

System Architecture

Frontend Architecture

  • Framework: React with TypeScript
  • Styling: Tailwind CSS with shadcn/ui components
  • State Management: TanStack React Query for server state management
  • Routing: Wouter for client-side routing
  • Build Tool: Vite for development and production builds

Backend Architecture

  • Runtime: Node.js with Express.js
  • Language: TypeScript with ESM modules
  • Database ORM: Drizzle ORM for type-safe database operations
  • Database: PostgreSQL (configured for Neon Database)
  • API Pattern: RESTful API with JSON responses

Project Structure

  • Monorepo Layout: Client, server, and shared code in separate directories
  • Client: React frontend located in /client
  • Server: Express backend located in /server
  • Shared: Common types and schemas in /shared

Key Components

Database Schema

The application uses PostgreSQL with the following main entities:

  • Documents: Legal documents with metadata (court filings, flight logs, etc.)
  • Names: Individuals mentioned in documents with mention counts
  • Connections: Relationships between individuals based on document co-mentions
  • Timeline Events: Chronological events related to documents and cases
  • Document-Name Relations: Many-to-many relationship tracking

Frontend Components

  • Dashboard: Main application interface with statistics and navigation
  • Document Import Center: Step-by-step guided document import process
  • AI-Powered Insights: Advanced document analysis and pattern recognition using Gemini AI
  • Document Management: Search, filter, and view legal documents
  • Network Visualization: Interactive graph showing relationships between entities
  • Timeline View: Chronological display of legal events
  • Name Analysis: Analysis of individuals mentioned across documents
  • Settings Page: User configuration for AI API keys and preferences
  • Export Functionality: Data export in multiple formats (CSV, JSON, PDF)

API Endpoints

  • Dashboard Stats: /api/dashboard/stats
  • Document CRUD: /api/documents
  • Name Analysis: /api/names
  • Network Data: /api/network
  • Timeline Events: /api/timeline
  • Export Functions: /api/export/{format}
  • Document Import: /api/documents/upload
  • Bulk Import: /api/documents/import-from-url
  • AI Status: /api/ai/status
  • AI Configuration: /api/ai/configure
  • AI Test: /api/ai/test
  • AI Document Analysis: /api/ai/analyze-document
  • AI Pattern Recognition: /api/ai/find-patterns
  • AI Network Insights: /api/ai/generate-insights

Data Flow

  1. Document Ingestion: Documents are stored with metadata and content
  2. Name Extraction: Names are extracted from documents and tracked
  3. AI Analysis: Documents are analyzed using Gemini AI for entity extraction, sentiment analysis, and pattern recognition
  4. Relationship Building: Connections are established between co-mentioned individuals
  5. Timeline Creation: Events are chronologically organized
  6. Visualization: Data is presented through interactive components with AI-powered insights
  7. Export: Processed data can be exported in various formats

AI-Powered Features

Gemini AI Integration

The platform integrates with Google's Gemini AI to provide advanced document analysis capabilities:

  • Entity Recognition: Automatic extraction of people, organizations, and locations from legal documents
  • Sentiment Analysis: Document sentiment scoring with confidence levels
  • Pattern Recognition: Identification of recurring themes and connections across document collections
  • Network Analysis: AI-powered insights about relationship networks and potential risk factors
  • Document Summarization: Automatic generation of concise document summaries
  • Smart Recommendations: Contextual suggestions for further investigation

AI Configuration

  • User-Controlled: Users provide their own free Gemini API key
  • Privacy-Focused: API keys are stored securely and never logged
  • Optional: All AI features are optional and the platform works without AI
  • Free Tier: Utilizes Gemini's generous free tier for document analysis

AI Capabilities

  • Document Analysis: /api/ai/analyze-document - Extracts entities, topics, and sentiment
  • Pattern Finding: /api/ai/find-patterns - Identifies patterns across document collections
  • Network Insights: /api/ai/generate-insights - Provides insights about relationship networks
  • Real-time Processing: AI analysis integrated into document import workflow
  • Batch Processing: Support for analyzing multiple documents simultaneously

External Dependencies

Database

  • Neon Database: Serverless PostgreSQL database
  • Connection Pooling: Using Neon's connection pooling for performance

UI Components

  • shadcn/ui: Comprehensive component library built on Radix UI
  • Radix UI: Headless UI components for accessibility
  • Tailwind CSS: Utility-first CSS framework

Data Visualization

  • D3.js: For network graph visualization and interactive charts
  • TanStack React Query: For efficient data fetching and caching

Development Tools

  • Vite: Fast development server and build tool
  • TypeScript: Type safety across the entire application
  • ESBuild: Fast JavaScript bundler for production builds

Deployment Strategy

Development

  • Development Server: Vite dev server for frontend, tsx for backend
  • Hot Reload: Both client and server support hot reloading
  • Environment Variables: DATABASE_URL for database connection

Production

  • Build Process: Vite builds frontend, ESBuild bundles backend
  • Static Assets: Frontend built to /dist/public
  • Server Bundle: Backend bundled to /dist/index.js
  • Database Migration: Drizzle migrations in /migrations

Environment Configuration

  • Development: NODE_ENV=development with local database
  • Production: NODE_ENV=production with production database
  • Database: PostgreSQL via DATABASE_URL environment variable

The application is designed to be deployed on platforms that support Node.js applications with PostgreSQL databases, with particular optimization for Replit's environment.