Python Bridge - Emotional Orchestration Engine (EOE)

The Python Bridge provides real-time, low-latency (<50ms) sensory processing for live streaming operations.

Purpose

While n8n handles asynchronous workflows, the Python Bridge handles:

• Real-time chat velocity monitoring
• Audio level detection
• OBS scene control
• Flash clipping triggers
• Transparent overlay rendering

Architecture

┌─────────────────────────────────────────────────┐
│          Python Bridge (EOE)                    │
├─────────────────────────────────────────────────┤
│                                                 │
│  ┌──────────────┐    ┌──────────────┐          │
│  │ State        │───▶│ OBS Control  │          │
│  │ Evaluator    │    │ (WebSocket)  │          │
│  └──────────────┘    └──────────────┘          │
│         │                                       │
│         ▼                                       │
│  ┌──────────────┐    ┌──────────────┐          │
│  │ Chat Monitor │    │ Audio Monitor│          │
│  │ (Twitch IRC) │    │ (PyAudio)    │          │
│  └──────────────┘    └──────────────┘          │
│                                                 │
└─────────────────────────────────────────────────┘

Setup

1. Install Dependencies

   cd python-bridge
   python3 -m venv .venv
   source .venv/bin/activate
   pip install -r requirements.txt

2. Configure OBS

   • Enable WebSocket Server (Tools → WebSocket Server Settings)
   • Port: 4455
   • Password: (set in .env)

3. Run

   python eoe_engine.py

State Machine

State	Trigger	Action
HYPE PEAK	Chat Velocity > 20 AND Audio > -15dB	SaveReplayBuffer + Flash Clip
DEAD AIR	Chat Velocity < 5 AND Audio < -40dB	AI Conversation Starter
GETTING COOKED	Chat Velocity > 25 AND Audio < -35dB	Sarcastic Overlay

Files (To Be Created in Phase 1)

• eoe_engine.py - Main orchestration engine
• state_evaluator.py - Entropy calculation
• obs_controller.py - OBS WebSocket client
• chat_monitor.py - Twitch IRC listener
• audio_monitor.py - Audio level detection
• flash_clipper.py - Automated clip generation