How we shaped a new wearable layer of intelligence in 24 hours

In just 24 hours, our team of 2 product designers and 2 business strategists redefined how AI could move beyond screens and become an ambient, proactive intelligence woven into everyday life.

This work took place during the 2026 AI & Startup Design Hackathon, bringing together students from Parsons School of Design at The New School and the University of Arizona, where 14 teams and 44 participants were challenged to imagine products that could compete in the AI era and in the world of 2035.

The result was Iris - an agentic AI platform for smart glasses designed to deliver proactive, contextual assistance through an ambient interface woven into daily life.

Reframing the original idea

Early in the hackathon, our team explored BandJam, a mobile app concept proposed by the business strategists to help musicians match, meet, and play together. However, the idea still relied on traditional patterns of the screen-based interface.

At that point, as the product designer driving UX vision and concept direction, I proposed reframing the problem entirely. Instead of designing another app or dashboard, I asked a different question: What if intelligence could live around you rather than inside your phone?

That shift led to the core vision behind Iris - an agentic AI system that operates as an ambient interface layer through smart glasses, delivering contextual assistance without requiring users to constantly return to a screen.

Understanding “the now”

To understand why this happens, our team looked at 3 layers of today's technology landscape: hardware, AI assistants, and apps.

Insight 01

Wearables bring AI closer to the real world, but remain constrained by immature interfaces. The hardware is there, but the intelligence layer is still shallow.

Insight 02

AI Assistants are powerful but rely on explicit prompts. They react to requests but rarely understand context.

Insight 03

Most tools solve single tasks. Users rely on multiple apps to manage everyday activities.

concept

Our solution was Iris - an agentic AI platform for smart glasses that learns how you live and proactively assists you throughout the day.

Instead of acting like a chatbot on your face, waiting for prompts and responding in isolated moments, Iris functions as a discrete personalized agent that builds context over time, asks follow-up questions when needed, and creates personalized workflows for each user.

Contextual Navigation

Our user starts at the airport, moving fast and already mentally overloaded. Instead of stopping to search crowded boards, Iris instantly finds the gate and checks the timing for him.

When he lands, Iris helps him get to the right baggage claim carousel without adding one more layer of friction to an already tiring travel day

Design Principles

Peripheral Information

The Iris interface is organized into distinct visual zones so information appears based on urgency and relevance, while the center remains completely clear to avoid obstructing vision or creating distraction.

simple design System

We designed a minimal design system built on simple shapes, bold colors, and signal-based elements to ensure maximum visibility and efficient use of zones.

AI-based Accessibility

Iris analyzes the real-world background in the user’s field of view and dynamically adjusts opacity and color contrast to maintain optimal legibility and visual clarity.

impact

Iris shifts the paradigm from basic chat-based assistance to longitudinal AI that supports users throughout everyday life.

Our team believes this shift will define the next major category of intelligent interfaces, and Iris is designed to help shape it.

Lesson learned

Strong product vision beats visuals. When you step back from existing patterns and rethink the problem itself, better ideas emerge.

Team Iris at the Top

My team took home first place at the AI & Startup Design Hackathon hosted by Designpreneurs, bringing together students from Parsons School of Design and the University of Arizona.