Everyone is building agents, chatbots, and AI in the cloud. Cool. But how do you make AI work in the real world — on a device, at the edge, locally on a kiosk that doesn’t have time to run a round trip to a server cluster?

That's the problem we're putting in front of you.

Build AI that runs where the physical world actually happens: retail, restaurants, healthcare, logistics.

10 teams (2-4 per team). $6,000 prize pool. Designed for experienced developers (3+ years) who are ready to build and deploy production-grade AI solutions on real-world hardware across enterprise verticals.

This is for engineers who want to ship something real. Edge inference, computer vision, voice on actual hardware, running locally. We’re judging on a live demo, not a slide deck. If you've got 3+ years building ML systems and you're tired of demos that only work on a laptop, this is your week.

Mentors and AWS solutions architects on-site throughout. Good builds win. Great builds shape the future.

Applications are open. Ten teams, screened on the strength of your idea.

Evaluation Criteria

• 30% Enterprise Problem Depth: Does this solve a real, painful problem in the chosen vertical? Is the ROI clear?
• 25% Hardware Integration: How well does the solution interface with edge devices, sensors, or point-of-sale hardware?
• 25% AI Quality & Innovation: Sophistication of the ML models, latency optimization, and creative approach to inference.
• 20% Live Demo Readiness: Can you prove it works? Production-readiness and reliability during the final presentation.

Prize Allocation

• FIRST PLACE: $3,000
• SECOND PLACE: $2,000
• THIRD PLACE: $1,000