Vest Practices is a wearable training system for tradespeople built on AI glasses.
The glasses see what the trainee sees, hear what they ask, and talks back: naming the part in their hand in under a second, answering questions from a sourced parts catalog, and walking them through training modules step by step, hands-free.
Under the hood: a first-person camera feed, a detector we trained on our own footage, a vision model that reads part markings, voice activation, and a synthesized voice in the ear. Every session is stored as structured data - what the trainee saw, asked, and heard - so the system remembers what they've learned and how, and every hour of training becomes labeled data for the next model.
In our demo, a first-time trainee builds a fan-and-motor circuit while the glasses identify each part and guide each step by voice.
Factories run on components whose specs aren't public. We work with them to collect those specs and inject them into the glasses' context, so the trainer knows their parts, their procedures, and their hazards.
