Urban Pulse

We navigate cities every day, but rarely feel known by them. We pass the same park, the same café, the same plaza — and nothing remembers us back. Urban Pulse is a mobile application that changes that relationship. It gives public spaces a personality, a memory, and a voice — turning the places we inhabit into companions that recognize and grow with us over time.



The Role of AI



Urban Pulse uses generative AI as its core design material, not as a feature layer. Every location in the app is assigned a distinct AI-generated personality shaped by its history, architecture, and community character. This personality remains consistent over time, while an ambient mood layer shifts daily in response to weather, events, and seasonal context — making each place feel alive without feeling arbitrary.



The application's most significant AI contribution is what we call situational memory. Rather than treating each user interaction as isolated, the system accumulates context across visits — what the user saved, what they said, what time of day they tend to arrive. When a user returns, the AI draws on this history to open a conversation that feels continuous, not generic. The city picks up where it left off.



A bond-depth model tracks the evolving relationship between user and place across four stages — Introduced, Familiar, Regular, and Kept — surfacing richer content and more personal exchanges as the relationship matures. An emotional map visualizes this spatially, showing which parts of the city a user truly knows versus which remain undiscovered. An AI memory system then weaves individual capsules — voice notes, photos, shared moments — into a living record that connects users to the community of people who have loved the same places.



Design Intent



Urban Pulse proposes that AI's most meaningful role in everyday life may not be productivity or automation, but recognition — the feeling of being remembered by the world around you. This project explores what it looks like when machine learning is used not to optimize behavior, but to deepen belonging.