🎯 Poker Video Processor

Automatically process poker videos from Airtable - downloads, transcribes, extracts insights, and updates records.

---

🚀 What It Does

1. Polls Airtable every 5 minutes for videos with Status = "Raw"
2. Downloads videos from YouTube, Google Drive, or direct URLs
3. Extracts audio using ffmpeg
4. Transcribes using OpenAI Whisper
5. Extracts insights using GPT-4:
   • Key quotes (tweetable)
   • Core poker philosophy
6. Updates Airtable with results and sets Status = "Extracted"

---

📋 Prerequisites

• Python 3.11+
• ffmpeg (for audio extraction)
• Airtable Personal Access Token (with read/write access)
• OpenAI API Key (for transcription and insights)

---

🛠️ Local Setup

1. Install Dependencies

# Create virtual environment
python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install packages
pip install -r requirements.txt

2. Configure Environment

Copy .env.example to .env and fill in your credentials:

cp .env.example .env

Edit .env:

AIRTABLE_API_KEY=your_personal_access_token
AIRTABLE_BASE_ID=appd81rBXhVWHn2xu
AIRTABLE_TABLE_ID=tblCnNsHMyGjXCXL6
OPENAI_API_KEY=your_openai_api_key
POLL_INTERVAL_SECONDS=300

3. Run the Service

Continuous mode (runs forever, checking every 5 minutes):

cd src
python main.py

One-shot mode (process once and exit):

cd src
python main.py --once

---

☁️ Deploy to Railway

Railway is a simple platform for hosting Python services. Free tier available.

1. Install Railway CLI

npm install -g @railway/cli

2. Login and Initialize

railway login
railway init

3. Set Environment Variables

railway variables set AIRTABLE_API_KEY=your_token
railway variables set AIRTABLE_BASE_ID=appd81rBXhVWHn2xu
railway variables set AIRTABLE_TABLE_ID=tblCnNsHMyGjXCXL6
railway variables set OPENAI_API_KEY=your_key
railway variables set POLL_INTERVAL_SECONDS=300

4. Create Procfile

Create a file named Procfile in the root directory:

worker: cd src && python main.py

5. Deploy

railway up

Your service will now run 24/7 in the cloud!

---

🎬 How to Use

Add a Video to Process

1. Go to your Airtable "Master Content Library"
2. Create a new record:
   • Content Title: "My Poker Strategy Video"
   • Source File/Link: https://youtube.com/watch?v=...
   • Status: "Raw"
   • Content Type: "Video"
3. Wait 5 minutes (or run python main.py --once)
4. Check the record - it will be updated with:
   • Transcription in "Core Philosophy" field
   • Key quotes in "Key Quotes" field
   • Status changed to "Extracted"

---

📁 Project Structure

poker-video-processor/
├── src/
│   ├── main.py              # Main service orchestrator
│   ├── video_processor.py   # Video download, transcription, insights
│   └── airtable_client.py   # Airtable integration
├── downloads/               # Temporary video/audio storage
├── logs/                    # Service logs
├── .env                     # Your credentials (DO NOT COMMIT)
├── .env.example            # Template for credentials
├── requirements.txt        # Python dependencies
└── README.md              # This file

---

🔧 Troubleshooting

"No module named 'yt_dlp'"

Make sure you're in the virtual environment:

source venv/bin/activate

"ffmpeg not found"

Install ffmpeg:

• Mac: brew install ffmpeg
• Ubuntu: sudo apt install ffmpeg
• Windows: Download from https://ffmpeg.org/

"Airtable API Error"

• Check your Personal Access Token has correct scopes:
  • data.records:read
  • data.records:write
  • schema.bases:read
• Verify Base ID and Table ID are correct

Videos Not Processing

• Check logs in logs/processor.log
• Verify video URL is accessible
• Make sure Status is exactly "Raw" (case-sensitive)

---

🎨 Customization

Change Poll Interval

Edit .env:

POLL_INTERVAL_SECONDS=600  # Check every 10 minutes

Modify Insight Extraction

Edit src/video_processor.py, method _extract_insights():

• Change the prompt
• Add more fields
• Use different AI models

Add More Status Values

Edit src/airtable_client.py to handle additional statuses like "Published", "Monetized", etc.

---

💰 Cost Estimates

OpenAI API Costs (per video):

• Whisper transcription: ~$0.006 per minute of audio
• GPT-4 Mini insights: ~$0.01 per video
• Total: ~$0.05 per 5-minute video

Railway Hosting:

• Free tier: $5 credit/month (plenty for this service)
• Paid: $5/month for 24/7 uptime

---

🚦 Next Steps

1. Test locally with one video
2. Deploy to Railway for 24/7 automation
3. Build Workflow 2: Content Derivative Generator
4. Add social media distribution
5. Track performance metrics

---

📞 Support

Check logs first:

tail -f logs/processor.log

Common issues are usually:

• Missing credentials
• Wrong Airtable field names
• Video URL not accessible

---

Built with ❤️ for poker content creators