3rd Place
MODO
AI Automation workflow powered by visual and audio Will put demo vid in github readme, youtube not letting me upload
Core trick
Runs YOLOv8 ONNX detection on camera frames, then sends the detections to Gemini, which decides which automation to trigger.
Description
AI Automation workflow powered by visual and audio
Will put demo vid in github readme, youtube not letting me upload
README excerpt
# Camera Copilot MVP # DEMO: https://youtu.be/W_Eu33I7SAQ A visual automation system using **YOLOv8 ONNX** for object detection and **Gemini AI** for semantic analysis. ## 🚨 IMPORTANT: Starting the App This app requires **TWO servers** to run: ### 1. Express Server (Port 5000) ✅ Auto-starts - Already running via the "Start application" workflow - Frontend + API endpoints ### 2. Python AI Backend (Port 8001) ⚠️ Manual start required - Handles **YOLOv8 object detection** with ONNX Runtime - **You MUST start this manually in a Shell tab!** ## 🚀 Quick Start ### Step 1: Start Python Backend Open a **Shell** tab and run: ```bash cd python_backend && python main.py ``` **What you'll see:** ``` 📥 Downloading YOLOv8n model from Hugging Face... (first time only) ✓ Model downloaded successfully (11.6MB) ✓ YOLOv8 model loaded successfully with ONNX Runtime INFO: Uvicorn running on http://0.0.0.0:8001 ``` **Keep this Shell tab open!** The Python backend will run in this terminal. ### Step 2: Use the App Once you see `"Uvicorn running on http://0.0.0.0:8001"`, your app is ready! Go to the **Flow Editor** page and click **"Run Flow"** to test the complete pipeline. --- ## 📝 Alternative: Set Up Auto-Start (Recommended) To automatically start both servers together: 1. Press `Cmd/Ctrl + K` → Search for "Workflows" 2. Click **"Create new workflow"** 3. Configure: - **Name**: `Python AI Backend` - **Command**: `bash start-python-backend.sh` - **Wait for port**: `8001