A powerful Windows desktop productivity and planning application built with Electron, featuring AI-powered smart scheduling, calendar management, app usage tracking, and intelligent workflow optimization.
- 📊 Product Requirements Document (PRD) - Complete feature specifications, technical architecture, and development roadmap
- ⚡ Quick Implementation Guide - Step-by-step setup instructions to get started immediately
- 🧠 Smart Scheduling Guide - AI-powered event scheduling documentation
- ✅ Integration Checklist - Smart scheduling testing and validation guide
- Intelligent Event Classification: LangChain + OpenAI automatically categorizes events as business, hobby, or personal
- Context-Aware Time Suggestions: AI suggests optimal time slots based on event type and existing schedule
- Business Hours Intelligence: Automatically schedules work events during 9-5 weekdays
- Work-Life Balance: Suggests hobby activities for evenings and weekends
- Conflict Detection: Avoids scheduling conflicts with existing events
- Local Fallback: Keyword-based classification when offline or API unavailable
- One-Click Scheduling: Select AI suggestions instantly with priority indicators
- Create events with title, description, time, and attendees
- Recurring events with flexible patterns
- Multiple calendar views (month, week, day, agenda)
- Drag-and-drop event scheduling
- Enhanced Event Modal: Modern Material-UI interface with smart suggestions panel
- Local Storage: All events stored locally for privacy and offline access
- Supabase-powered user authentication
- OAuth integration (Google, Microsoft)
- Secure session management
- OAuth 2.0 Gmail connectivity
- Smart email-to-event conversion
- Meeting invitation parsing
- Conflict detection and resolution
- Native Windows notifications
- Customizable reminder timing
- Snooze functionality
- Quick actions (reschedule, complete)
- App usage tracking (privacy-focused)
- Time spent analysis per application
- Productivity insights and reports
- Focus time optimization
- Windows App Tracking: Monitors active applications using PowerShell (no admin required)
- Auto-start at Windows boot
- System Tray Integration: Minimize to tray with quick access popup
- Tray Popup: Compact interface for quick event creation and viewing
- Global keyboard shortcuts
- Note: System tray functionality has been tested on Windows 11
- Frontend: Electron + React + TypeScript
- Backend: Supabase (PostgreSQL + Auth + Realtime)
- AI/ML: LangChain + OpenAI GPT-3.5-turbo for smart scheduling
- UI: Material-UI + MUI X Date/Time Pickers
- State Management: Redux Toolkit
- Calendar: React Big Calendar
- Notifications: node-notifier
- Windows APIs: Native app tracking via PowerShell
- Build System: Electron Forge with Webpack
-
Review the documentation:
- Read the PRD for complete feature specifications
- Follow the Implementation Guide for setup
- Check the Smart Scheduling Guide for AI features
-
Initialize the project:
npm create electron-app@latest FlowGenius -- --template=typescript-webpack cd FlowGenius -
Install dependencies:
# Core dependencies npm install react react-dom @types/react @types/react-dom npm install @supabase/supabase-js @mui/material @emotion/react @emotion/styled npm install @reduxjs/toolkit react-redux node-notifier dayjs react-big-calendar # AI/Smart Scheduling dependencies npm install langchain @langchain/openai @langchain/core npm install @mui/x-date-pickers @mui/icons-material
-
Set up environment variables:
# Create .env file with: OPENAI_API_KEY=your_openai_api_key_here SUPABASE_URL=your_supabase_project_url SUPABASE_ANON_KEY=your_supabase_anon_key -
Set up Supabase:
- Create project at supabase.com
- Run the database schema from the implementation guide
- Configure environment variables
-
Start development:
npm run dev
npm run dev# Clean previous builds
npm run clean
# Build Windows installer
npm run make:winThe build process creates two installer types in out/make/squirrel.windows/x64/:
FlowGenius-Setup.exe- Standard Windows installerFlowGeniusSetup.msi- MSI installer for enterprise deployment
After installation, FlowGenius will:
- Install to
%LOCALAPPDATA%\FlowGenius\ - Add itself to Windows Start menu
- Be searchable in Windows search
- Start with Windows (if enabled in settings)
The AI automatically analyzes event titles and descriptions to classify them:
- Business Events: Meetings, calls, presentations, client work
- Suggests 9 AM - 5 PM weekday slots
- Hobby Events: Sports, music, gaming, creative activities
- Suggests evening (6-10 PM) and weekend slots
- Personal Events: Appointments, errands, family time
- Flexible timing suggestions
- 85-95% Classification Accuracy for clear event types
- Conflict Avoidance - Never suggests overlapping times
- Priority Ranking - Optimal, Good, Fair, Available time slots
- Cost Effective - ~$0.002 per classification via OpenAI API
- Reliable Fallback - Local keyword matching when offline
- One-Click Scheduling: Select suggested times instantly
- Visual Indicators: Clear classification with confidence levels
- Manual Override: Always maintain full control over scheduling
- Professional UI: Modern Material-UI design with priority badges
FlowGenius runs in the background with a system tray icon for quick access:
- Minimize to Tray: Close button minimizes to tray instead of exiting
- Quick Access Popup: Right-click tray icon for compact interface
- Event Creation: Create events directly from the tray popup
- Today's Schedule: View upcoming events at a glance
- Productivity Stats: See app usage time in real-time
- Windows 11 Tested: System tray features have been verified on Windows 11
The productivity tracking feature monitors your app usage to provide insights:
- No Admin Required: Uses PowerShell to detect active applications
- Privacy Focused: All tracking data stays local
- Real-time Updates: Updates every 5 seconds
- Accurate Detection: Tracks window titles and process names
- Resource Efficient: Minimal CPU and memory usage
- Fixed icon not appearing in system tray
- Resolved production build icon path issues
- Added proper resource packaging in build configuration
- Simplified PowerShell implementation for better reliability
- Removed admin requirement for app detection
- Added fallback methods for window detection
- Improved error handling and logging
- Streamlined tray popup interface
- Fixed event saving from tray popup
- Improved location field handling in events
- Removed debug buttons from production build
- Enhanced AI assistant header sizing
- Fixed webpack configuration for proper production builds
- Resolved "require is not defined" errors
- Added MSI installer generation
- Improved Windows integration for app search
| Phase | Duration | Focus |
|---|---|---|
| Phase 1 | 2 weeks | Project Setup + Authentication |
| Phase 2 | 2 weeks | Core Calendar Functionality |
| Phase 3 | 2 weeks | AI Smart Scheduling + LangChain Integration |
| Phase 4 | 2 weeks | Notifications + Gmail Integration |
| Phase 5 | 2 weeks | App Usage Tracking + Analytics |
| Phase 6 | 2 weeks | Polish + Testing + Deployment |
Total Development Time: 12 weeks
- Performance: < 3s cold start, < 500ms view switching, < 2s AI classification
- AI Accuracy: 90-95% business events, 85-90% hobby events
- User Adoption: 70% suggestion acceptance rate
- Time Savings: 40% reduction in manual scheduling time
- Reliability: < 1% crash rate
- Security: Encrypted data, secure authentication
FlowGenius is designed with privacy-first principles:
- Local Storage Only: All calendar events and data stored locally on your device
- No Cloud Sync: Events are not synced to Google Calendar or other cloud services
- App usage tracking is local-only by default
- AI Processing: Event titles/descriptions sent to OpenAI for classification only
- Local Fallback: Keyword matching works entirely offline
- No Content Capture: No keystrokes or sensitive data captured
- All data encrypted at rest and in transit
- User consent required for all tracking and AI features
- GDPR compliant data handling
- Cost: ~$0.002 per event classification
- Monthly Estimate: $1-5 for typical usage (50-250 events)
- Response Time: < 2 seconds for AI classification
- Fallback: Free local classification when API unavailable
- AI Classification: < 2 seconds with OpenAI
- Local Classification: < 100ms with keywords
- Time Slot Generation: < 1 second
- UI Responsiveness: Real-time suggestions as you type
- Primary: Windows 10/11 (64-bit)
- Future: macOS and Linux support planned
This is a personal productivity application, but the architecture and implementation can serve as a reference for similar projects. Key learnings include:
- Electron + React + TypeScript best practices
- Supabase integration patterns
- LangChain + OpenAI integration in desktop apps
- AI-powered UX design patterns
- Windows system integration
- Privacy-focused analytics implementation
├── README.md # This file - project overview
├── PRD_FlowGenius.md # Complete product requirements
├── IMPLEMENTATION_GUIDE.md # Step-by-step setup guide
└── FlowGenius/
├── SMART_SCHEDULING_GUIDE.md # AI scheduling implementation guide
├── INTEGRATION_CHECKLIST.md # Smart scheduling testing guide
└── SMART_SCHEDULING_INTEGRATION_COMPLETE.md # Feature completion summary
FlowGenius follows these core principles:
- Intelligence: AI-powered insights without complexity
- Simplicity: Clean, intuitive interface
- Performance: Fast, responsive, lightweight
- Privacy: User data stays secure and private
- Integration: Seamless workflow with existing tools
- Reliability: Graceful fallbacks when AI services are unavailable
Ready to experience AI-powered productivity? Start with the Implementation Guide and build your intelligent productivity powerhouse! 🧠💪