WeMakeDevs Hackathon 2026 🎮 Demo
Connecting…
🎬 Upload Video
Upload a short video (10–60 s recommended). The agent will extract frames, transcribe audio, and detect objects.
Drag & drop your video here
or click to browse — MP4, MOV, WebM supported
Uploading… 0%
🧠 Generate AI Study Notes
Uses an LLM (OpenAI) to turn transcript + detections into structured notes, formulas, and viva questions. Requires running Full Analysis first.
⚠️ No analysis found yet. Go to "Upload & Analyze" tab and run Full Analysis first.
📡 Real-Time Vision Agent Stream
Stream from your webcam or screen — each 2s chunk gets instant YOLO detection with bounding boxes, pose estimation, fitness coaching, transcription, and agent reasoning. Supports 7 exercises with rep counting, streak tracking, and voice feedback.
0
Chunks
Avg Latency
0
Objects
0
Frames
P90 Latency
Model FPS
00:00
Camera
🏋️ Live Coach
0
reps
score —
Start streaming to get live form feedback.
Waiting to start…
🤖 Agent Response
Click any label above to ask the agent…
🌐 Ingest Video from URL
Paste a public video URL (YouTube, Vimeo, Facebook, X/Twitter, Twitch) and we'll download, analyze, and generate notes automatically.
📋 Supported: YouTube, Vimeo, Facebook, X/Twitter, Twitch, Dailymotion, Instagram
⚡ If captions are available, we use those (faster & free). Otherwise we download and transcribe.
Downloading…
🤖 Ask the Vision Agent
Ask anything about your analyzed video. The agent uses a 2-tier intelligence system: an instant FastReply (under 500ms) from YOLO detections + transcript, followed by a deep LLM-powered PolishReply with full context analysis. Powered by Gemini/OpenAI/Cloudflare cascade.
⚠️ No analysis found yet. Upload & analyze a video first, then come back to chat.
💬
Start a conversation
Ask about objects detected, transcript content, or any scene details
🧪 Interactive Quiz
Generate MCQ and short-answer questions from your video analysis. Test your understanding of the content!
⚠️ No analysis found yet. Upload & analyze a video first, then generate a quiz.