Prototype walkthrough
AI-Driven Gardening Platform
Problem
People want to grow their own food, but the knowledge and tools are scattered, and general AI chatbots forget your garden between chats and never remind you when to act.
Outcome
A mobile prototype tested with 30+ people, shaped by research with Brooklyn community gardens, with 60+ waitlist sign-ups.

Interviewed gardeners at Brooklyn community gardens.
User research
Built a clickable mobile prototype.
UI/UX design, prototyping
Usability tested with 30+ people.
Usability testing
Reworked planning, care and community from the findings.
Iteration
What testing the first prototype showed
“Everybody forgets to water.”
Care reminders were the most valued feature.
“Why isn't my tomato flowering?”
Troubleshooting was the most-wanted missing feature.
Gardens aren't grids
Mixed sun and irregular plots broke the fixed grid.
→ Planning moved from a grid to a list
Local beats big-box
Gardeners preferred local sources for purchasing supplies.
→ Marketplace removed for now





30+
individuals in usability testing
(homeowners and community gardeners)
60+
waitlist sign-ups
Dec 2025
selected for and presented at Cornell Tech Presents: Builders & Disruptors in the AI Era (event recap)
40 pixel-art plant icons I created with Midjourney for our plant dataset, which we researched for New York State and Brooklyn growing conditions, including native plants.
