Inspiration:
Coming to a new city as an international student is a unique kind of overwhelming. You’re surrounded by incredible architecture and historic piers, but without a guide, they are just cold shapes of glass and concrete. We wanted to bridge the gap between being an outsider and being an insider.
"Been there, Done That" was born from our personal journey as international students in Canada. We felt a constant urge to "know" our new home, but typical tour guides felt too formal. We wanted to build something that didn't just "see" a building, but understood its "lore" the hidden history and fun facts that usually take years of living in a city to discover.
What it does
"Been there, Done That" is a real-time discovery engine that turns your camera into a storyteller.
- Instant Identification: Point your camera at any landmark, and our AI identifies it using live visual grounding.
- The Lore Drop: Instead of dry Wikipedia facts, the app serves up "Lore", concise, engaging stories and "Did You Know?" facts.
How we built it
- The project is built on a modern React Native stack using Expo.
- We integrated Gemini 2.5 Flash to handle multimodal processing.
- We utilized expo-camera for real-time snapshots and expo-location for reverse geocoding.
Challenges we ran into
The biggest hurdle was API Resilience. During development, we encountered several "Model Not Found" errors as we navigated the transition between different Gemini API versions. We learned the hard way that in a fast-moving AI landscape, you can't just hardcode an endpoint; you have to build diagnostic tools to query the server for available models.
Log in or sign up for Devpost to join the conversation.