Inteligência Cerimonial

Large events share a known bottleneck: recognizing who just arrived, in the moment, without a queue or a clipboard.

The exploration

We prototyped a real-time facial recognition system for guest arrival:

• Live video in the browser via WebRTC (aiortc), nothing to install at the reception desk;
• Facial recognition service in Python, packaged in Docker;
• A login-gated app for the event team and a Chrome extension to operate on top of the tools they already use;
• Stack: Python, Flask, aiortc, Docker and JavaScript.

Status: an R&D proof of concept. It validated technical feasibility and was archived.

Why we show this

Real-time computer vision is rare muscle. The prototype went to the archive; the capability stays in the arsenal.