Say: "Send the report to the team and schedule a review for tomorrow" Saydo: β Email sent, β Calendar event created, β Slack notified, β Task added
- π€ Join meetings automatically with an AI bot that listens and learns
- π― Extract actions like scheduling follow-ups, sending emails, or updating spreadsheets
- π Execute tasks across Google Workspace and 100+ business tools automatically
- π Learn & improve from every conversation to better understand your workflow
- π Scale infinitely without hiring a single administrative assistant
That's Saydo. Not just another transcription toolβit's your first AI employee.
Real Meeting Transcript:
Cuzin: "Let's set aside some time on Thursday at 12 pm for a quick sync on the Q3 roadmap"
Magdy: "Wednesday at 12:00 works, I'll send out the invite shortly"
Saydo AI Generated:
β
Meeting Summary: Planning Q3 roadmap sync scheduled for Thursday at 12 pm
β
Calendar Event: "Q3 Roadmap Sync" created for Thursday 12:00 PM
β
Todo Created: "Run updates" assigned to Magdy
β
Smart Context: AI understood "Thursday" means next Thursday from meeting date
All actions pending your one-click approval!
- 73% of professionals spend 2+ hours weekly on meeting follow-ups
- $37 billion lost annually due to inefficient meeting management
- Current solutions stop at transcription or require complex integrations
|
Deploy intelligent bots that join your meetings, understand context, and extract actionable insights.
|
Review and execute AI-suggested actions with a single tap. From emails to calendar events.
|
Get your first AI meeting assistant running in under 5 minutes:
# Clone the repository
git clone https://github.com/saydo-ai/callops.git
cd callops
# Install client dependencies
cd client && npm install
# Install server dependencies
cd ../server && pip install uv && uv sync
# Set up environment variables
cp .env.example .env
# Add your API keys (OpenAI, Deepgram, Google OAuth, Firebase)
# Start the platform
# Terminal 1 - Client
cd client && npm run dev
# Terminal 2 - Server
cd server && uv run python -m app.main
# Open http://localhost:3000 and experience the magic! πBuilt with Next.js 15 and TypeScript, our frontend makes AI meeting management intuitive:
- π± Mobile-First Design: Optimized for on-the-go professionals
- β‘ Real-time Updates: Live transcription and action generation
- π― Smart Actions: Color-coded, one-click executable tasks
- π Secure Auth: Firebase authentication with Google OAuth
- Meeting Dashboard: Chronological view with smart filtering
- AI Agent Settings: Customize your AI's capabilities
- Voice Commands: Quick access to recording modes
- Action Review: Approve or reject AI suggestions
Powered by FastAPI and cutting-edge AI services:
- ποΈ Advanced Voice Processing: Deepgram for 95%+ accuracy transcription
- π§ GPT-4 Intelligence: Context-aware action extraction
- π Google Workspace Integration: Calendar, Gmail, Sheets, Tasks
- π Real-time Processing: WebSocket-based live updates
-
Meeting Bot System (
/server/app/services/)- Cloud-based bots via Recall.ai
- Automatic meeting joining
- Real-time transcript processing
- Speaker identification
-
AI Agent (
/server/app/services/openai_agent.py)- Dynamic capability loading
- Function calling for actions
- Context-aware processing
- Human-in-the-loop approval
-
Action Executor (
/server/app/services/action_executor.py)- Google Calendar events
- Gmail composition
- Sheets updates
- Task creation
- Document generation
- Automatic action item extraction
- Follow-up email drafts
- Calendar event scheduling
- Meeting summary generation
- Task assignment tracking
- CRM updates (coming soon)
- Proposal generation
- Meeting notes documentation
- Next steps automation
- Deal tracking
- Attendance tracking
- Key points extraction
- Resource sharing automation
- Follow-up sequences
- Quiz generation (coming soon)
- Lead qualification notes
- Opportunity updates
- Follow-up scheduling
- Pipeline management
- Activity logging
| Frontend | Backend | AI/Voice | Infrastructure |
|---|---|---|---|
| Next.js 15 | FastAPI | OpenAI GPT-4 | Firebase |
| React 19 | Python 3.12+ | Deepgram | Google Cloud |
| TypeScript | Pydantic | Recall.ai | WebSockets |
| Tailwind CSS | AsyncIO | Speaker Diarization | Docker |
| GraphQL |
"We saved 15 hours per week on meeting follow-ups. It's like having a super-efficient assistant who never misses a detail."
β Sarah Chen, Product Manager
"The AI caught that Cuzin wanted Thursday but I said Wednesday - it created both options for us to choose. That's real intelligence."
β Magdy, Engineering Lead
"Our sales team closes 30% more deals because follow-ups happen instantly. No more dropped balls."
β Mike Johnson, Sales Director
"The AI understands context better than most humans. It knows when 'next week' means and who should be invited."
β Lisa Park, Operations Manager
- Google Meet bot integration
- Real-time transcription with speaker identification
- AI-powered action extraction
- Google Workspace integration (Calendar, Gmail, Sheets)
- Mobile-responsive design
- One-click action execution
- Zoom integration
- Microsoft Teams support
- Slack integration
- Custom AI training
- Voice command activation
- Salesforce CRM connector
- Advanced analytics dashboard
- Team collaboration features
We love contributors! Here's how you can help:
- Fork the repository
- Create your feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
- Add new action types (see Adding New Actions)
- Improve UI/UX components
- Add more AI capabilities
- Write tests
- Improve documentation
- Report bugs
- REST API Reference - Complete API endpoint documentation
- Google Workspace Setup Guide - Step-by-step integration setup
- Adding New Actions - Developer guide for extending actions
- GraphQL Schema - Data model definitions
- All data encrypted at rest and in transit
- OAuth 2.0 for third-party integrations
- User-scoped data access
- No automatic action execution without approval
- Regular security audits
This project is licensed under the MIT License - see the LICENSE file for details.
Built with β€οΈ by the RINDA team and amazing contributors worldwide.
Special thanks to:
- Recall.ai for enterprise-grade meeting bot infrastructure
- Deepgram for incredible speech recognition
- OpenAI for groundbreaking AI models
- Our beta users for invaluable feedback


