AI Local Discovery
Defining how AI helps people discover the physical world on Meta Ray-Ban Display Glasses with Neural Band.
“How can AI understand where you are, what you’re looking at, and what you want, before you ask?”
- Role
- Product Designer
- Timeline
- 2022 – 2023
- Collaborators
- 1 Product Designer · Product Manager · AI/ML Engineers · UX Research · Content Design · Engineering
- Related public release
- Meta AI Glasses – GPS Navigation
Impact
As one of two product designers driving Local Discovery, I helped evolve the experience from a traditional location search into a context-aware AI agent for Meta’s next-generation display glasses.
Working across Product, AI, Engineering, and Research, we expanded the product vision beyond navigation to explore how AI could understand user intent through location, surroundings, and context. I helped define core AI use cases, design end-to-end interaction flows, facilitate cross-functional strategy workshops, and build functional prototypes that influenced the long-term direction of AI experiences on the platform.
The Challenge
Search is reactive. AI has the opportunity to be proactive.
Rather than asking users to search for nearby places, we explored how AI could recognize context, anticipate intent, and surface relevant information naturally within the flow of everyday life.
The challenge wasn’t designing another maps experience - it was defining how an AI agent should observe, reason, and assist within the physical world.
Designing the Experience
Instead of adapting mobile patterns, we designed interactions specifically for wearable AI. Together, we explored how AI could:
- Understand location, surroundings, and user intent to deliver contextual recommendations.
- Blend conversational voice interactions with glanceable visual feedback.
- Seamlessly connect local discovery with navigation and follow-up actions.
- Leverage predictive intent and generative UI to proactively assist users.
- Transition naturally between lightweight wearable interactions and richer phone experiences.
To validate these concepts, I collaborated closely with Engineering to build functional AI prototypes using LLaMA and partnered with Product and Research to facilitate workshops that aligned the organization around a shared product vision.
Design Principles
Intent over search
The best AI experiences understand what users are trying to accomplish - not simply what they ask.
Context over commands
Location, surroundings, and timing provide valuable signals that reduce the need for explicit input.
Assistance over interruption
AI should surface information when it’s helpful, while remaining unobtrusive and allowing users to stay present in the real world.
Behind the scenes
Reflection
This project fundamentally changed how I think about AI product design. Building intelligent products isn’t about replacing search with conversation - it’s about designing systems that perceive context, reason about intent, and deliver assistance that feels both timely and effortless.