🎙️ Meeting Recorder - Tamil/English Transcription

Dual API system: OpenAI Whisper + DeepSeek/OpenAI Summarization for maximum cost efficiency

✨ Features

• 🌍 Direct English Transcription - Speak Tamil, get accurate English transcripts
• 🧹 Post-processing Cleanup - Removes artifacts, repetitions, and trailing dots
• 📝 Dual API Support - OpenAI for transcription, DeepSeek/OpenAI for summaries
• 💰 Cost Optimization - DeepSeek is 90-95% cheaper than OpenAI for summaries
• 🔄 API Switching - Switch between providers dynamically
• 💻 Web App - Upload audio files via browser
• 🍎 Mac App - Menu bar recording with manual summary generation
• 📥 Download History - Access all transcriptions and summaries
• 🎯 Real-time Processing - Fast transcription with OpenAI Whisper

---

💰 Cost Comparison: Your $2 Budget

Provider	Input Cost	Output Cost	Summaries per $2	Savings
DeepSeek	$0.14/1M tokens	$0.28/1M tokens	~3,600	95% cheaper
OpenAI GPT-3.5	$0.50/1M tokens	$1.50/1M tokens	~1,300	Baseline
OpenAI GPT-4o	$2.50/1M tokens	$10.00/1M tokens	~200	Most expensive

🎉 DeepSeek gives you 18x more summaries for the same budget!

---

🚀 Quick Start

Prerequisites

# Check Python version (3.8+ required)
python --version

# Check if git is installed
git --version

1. Clone and Setup

# Clone the repository
git clone <your-repo-url>
cd Summary

# Create virtual environment
python -m venv venv

# Activate virtual environment
source venv/bin/activate  # On macOS/Linux
# or
venv\Scripts\activate     # On Windows

# Install dependencies
pip install -r requirements.txt

2. DUAL API SETUP (New!)

# Run the interactive setup script
python setup_dual_api.py

# Or manually configure your .env file:
# OPENAI_API_KEY=sk-proj-your-openai-key-here
# DEEPSEEK_API_KEY=sk-your-deepseek-key-here  
# API_PROVIDER=auto

What each option means:

• auto - Uses DeepSeek if available (cheapest), falls back to OpenAI
• deepseek - Forces DeepSeek for summaries (maximum savings)
• openai - Forces OpenAI for summaries (more expensive but familiar)

3. Start the Server

# Start the dual API server
python app.py

Expected Output:

🎙️ Meeting Recorder - DUAL API SYSTEM
📝 Features: Audio Upload → OpenAI Whisper → DeepSeek/OpenAI Summary
🤖 API Status:
   - OpenAI: ✅ Configured
   - DeepSeek: ✅ Configured
   - Active Provider: deepseek
💰 COST SAVINGS: Using DeepSeek for summaries (90-95% cheaper than OpenAI)
📊 Budget Estimate: $2 = ~3,600 summaries with DeepSeek vs ~200 with OpenAI
🌐 Server starting at: http://localhost:9000

---

🔄 API Provider Switching

Switch via API (Recommended)

# Switch to DeepSeek (cheapest)
curl -X POST -H "Content-Type: application/json" \
     -d '{"provider":"deepseek"}' \
     http://localhost:9000/api/switch-provider

# Switch to OpenAI
curl -X POST -H "Content-Type: application/json" \
     -d '{"provider":"openai"}' \
     http://localhost:9000/api/switch-provider

# Use auto-selection (recommended)
curl -X POST -H "Content-Type: application/json" \
     -d '{"provider":"auto"}' \
     http://localhost:9000/api/switch-provider

Switch via Environment Variable

# Set provider for current session
export API_PROVIDER=deepseek
python app.py

# Or update .env file
echo "API_PROVIDER=deepseek" >> .env

---

✅ Health Check Commands

Verify Server Status

# Check if server is running with dual API info
curl -s http://localhost:9000/health | python -m json.tool

Expected Response:

{
    "status": "healthy",
    "openai_configured": true,
    "deepseek_configured": true,
    "active_provider": "deepseek",
    "service": "Dual API Meeting Recorder - OpenAI Whisper + OpenAI/DeepSeek Summarization",
    "total_transcriptions": 46,
    "features": [
        "dual_api_support",
        "openai_whisper_transcription", 
        "deepseek_openai_summarization",
        "cost_optimization",
        "api_switching"
    ],
    "cost_savings": {
        "deepseek_vs_openai": "90-95% cheaper for summarization",
        "recommendation": "Use DeepSeek for summaries to maximize your $2 budget"
    }
}

Test API Connections

# Test both API connections
python setup_dual_api.py test

# Expected output:
# 🧪 Testing API Connections...
# ✅ OpenAI connection successful
# ✅ DeepSeek connection successful

Check Database

# Verify database includes API provider tracking
sqlite3 transcriptions.db "SELECT id, filename, api_provider, summary_provider, created_at FROM transcriptions ORDER BY created_at DESC LIMIT 5;"

---

🖥️ Usage Guide

Web App

1. Open Browser: Navigate to http://localhost:9000
2. Upload Audio: Choose your audio file (supports .m4a, .wav, .mp3)
3. Get Transcription: Automatic transcription with OpenAI Whisper
4. Generate Summary:
   • Uses DeepSeek by default (if configured) - 18x cheaper!
   • Falls back to OpenAI if DeepSeek unavailable
5. Download: Access transcription and summary files with provider info

Mac App

1. Install:

cd MeetingRecorderApp ./build.sh cp -R release/MeetingRecorder.app /Applications/

2. **Launch**:
```bash
open /Applications/MeetingRecorder.app

3. Use: Look for 🎤 icon in menu bar → Start Recording → Stop Recording → View History

---

🔧 Troubleshooting

API Configuration Issues

# Check which APIs are configured
python setup_dual_api.py test

# Reconfigure APIs
python setup_dual_api.py

# Check current provider
curl -s http://localhost:9000/health | grep active_provider

Cost Optimization

# Force DeepSeek for maximum savings
export API_PROVIDER=deepseek
python app.py

# Check summary provider in history
curl -s http://localhost:9000/history | grep summary_provider

Server Won't Start

# Check if port 9000 is in use
lsof -ti:9000

# Kill existing processes
pkill -f "python.*app.py"
lsof -ti:9000 | xargs kill -9

# Restart server
python app.py

OpenAI API Issues

# Verify API key is set
echo $OPENAI_API_KEY

# Test API key
curl -H "Authorization: Bearer $OPENAI_API_KEY" \
     -H "Content-Type: application/json" \
     https://api.openai.com/v1/models | head -20

DeepSeek API Issues

# Verify DeepSeek API key
echo $DEEPSEEK_API_KEY

# Test DeepSeek API
curl -H "Authorization: Bearer $DEEPSEEK_API_KEY" \
     -H "Content-Type: application/json" \
     https://api.deepseek.com/v1/models | head -20

---

🎯 Quick Reference

# Essential commands
python setup_dual_api.py                        # Configure dual APIs
python app.py                                    # Start server
curl -s http://localhost:9000/health            # Check health + API status
pkill -f "python.*app.py"                      # Stop server

# API switching
curl -X POST -H "Content-Type: application/json" -d '{"provider":"deepseek"}' http://localhost:9000/api/switch-provider
curl -X POST -H "Content-Type: application/json" -d '{"provider":"openai"}' http://localhost:9000/api/switch-provider

# Database with provider tracking
sqlite3 transcriptions.db "SELECT COUNT(*) FROM transcriptions WHERE summary_provider='deepseek';"

# Cost monitoring
sqlite3 transcriptions.db "SELECT summary_provider, COUNT(*) as count FROM transcriptions WHERE summary IS NOT NULL GROUP BY summary_provider;"

---

💡 Pro Tips for Maximizing Your $2 Budget

1. Use DeepSeek for summaries: Set API_PROVIDER=auto or deepseek
2. Monitor usage: Check /history endpoint to see which provider was used
3. Batch processing: Upload multiple audio files and summarize them all with DeepSeek
4. Test both providers: Compare quality and pick your preference

🎉 With DeepSeek, your $2 budget will last months instead of days, giving you 18x more meeting summaries!

---

🚀 You're all set! Your Meeting Recorder now supports both APIs with massive cost savings through DeepSeek integration.