Starrtea

Inspiration

Have you ever had an idea? Every trip starts the same way: a voice memo at 2am, a reel saved weeks ago, a screenshot of a restaurant you'll never find again. The inspiration is real. The chaos is realer. We didn't want to build another planner. Planners assume you already know what you want. We wanted to build something for the moment before that when you have a feeling but not a plan, a vibe but not an itinerary. That's Starrtea. The space between wanting to go somewhere and actually going.

What it does

Three steps. One swirl. Braindump. Throw everything onto a spatial canvas, photos, screenshots, voice memos, half-baked ideas, etc no structure required. No judgment. Just your chaos, in one place. Synthesize. One tap. Gemini reads the mess, finds the threads, and clusters your sparks into something coherent. That random screenshot? It's now sitting next to a jazz bar two blocks away in Kyoto. The chaos finds its own shape. Plan. A roadmap built around your taste, your pace, your trip, not an algorithm's best guess. Discover notes from people who actually went. Pick up wherever you left off.

How we built it

We used our own tools the way we hoped our users would. Ideas lived in Notability first... messy, spatial, non-linear. Mind maps next to sketches next to screenshots of competitors. It reminded us why the braindump matters: you can't find the good idea until you've put the bad ones somewhere safe. Gemini helped us make sense of our own research by clustering interview notes, stress-testing our "how might we" questions, and shaping the synthesize interaction at the core of the app. Getting that step right was the hardest design problem: AI that's powerful enough to help, but quiet enough that the trip still feels like yours. We both opened Figma for the first time at the start of this hackathon. We broke our design system twice. The third version held. By the end we had a full hi-fi system: midnight navy, sage tea, honey gold, warm cream. Our persona is Phoebe — early riser, culture chaser, planning something she's been putting off for months. Every screen, every word of microcopy, got checked against her first.

Challenges we ran into

Making AI feel like a collaborator instead of the main character. Gemini could plan the whole trip. That's not the point. The hard work was designing restraint, making the synthesis feel like a reveal, not a takeover. Spatial canvases sound freeing until you're trying to build one. The braindump screen went through more iterations than anything else before the dotted grid, drift behavior, and snap-points felt like play.

Accomplishments that we're proud of

The moment the swirl happens. We tested it with real people and got audible reactions from the volunteers.

What we learned

The transition is the product. "Chaos into clarity" isn't a tagline — it's the entire UX problem. Getting from braindump to roadmap has to feel like something, not just work. Constraints make you intentional. No Figma experience meant no shortcuts. Every choice was deliberate. We'd do it again. Designing the way your users work changes how you work. Following the thread instead of filling in the doc.

What's next for Starrtea

A real synthesize engine. Discovery Den firsthand traveler notes, curated cities, location memory, friend-following. Trip recaps that become scrapbooks. Postcards you can actually send. Social collaboration on mental inventory (brain dumps)