Personal Project
Leaf Me Alone
Plant parents shouldn't have to choose between traveling and keeping their plants alive. I designed and built a two-sided mobile marketplace to connect plant owners with local, trusted plant sitters.
A marketplace where plants don't have to suffer when you travel
I kept coming home from trips to struggling plants - drooping leaves, dry soil, the quiet guilt of a neglected pothos plant. I realized there was no Rover for houseplants, so I decided to build one.
Leaf Me Alone is a 0-to-1 mobile marketplace where plant owners can book trusted local sitters, share detailed care instructions per plant, and get photo check-ins throughout the trip. Sitters can list their availability, build a plant-focused profile, and choose flexible handoff options - from a single visit to a full multi-week sit.
I designed and built the whole thing myself using React Native and Expo - an excuse to work outside the constraints of a B2B product org, and to finally scratch an itch I'd had for years.
There's no Rover for plants - and it shows
There's no app built specifically for plant care. Nothing with plant-specific care notes, no way to communicate that your monstera needs indirect light and hates tap water, no trust signal that the person showing up actually knows the difference between over-watering and under-watering. The closest thing is a group chat and good luck.
I started talking to peers - city friends who travel often, with full apartments of plants they'd accumulated over the pandemic. We were all running the same workarounds: leaving verbose sticky notes, texting neighbors instructions we hoped they'd follow, or just accepting the loss. Nobody had found a real solution because no real solution existed.
The real gap
Plant owners had no trusted, plant-specific channel for care. And people who genuinely know and love plants had no way to offer that as a service.
The problem was two-sided: both the owner and the sitter were underserved. That made the design challenge genuinely interesting - and harder than I expected.
Designing for two people at once
Two-sided marketplaces are hard to design because you're always optimizing for two very different mental models in the same product. I started by mapping both journeys separately - what does an owner need to feel confident leaving their plants? What does a sitter need to feel set up for success before they ever walk through the door?
The core tension was trust. Plants are personal, sometimes sentimental, and highly variable in care needs. A fiddle-leaf fig is not a succulent. I designed around that: detailed per-plant care profiles, sitter profiles with plant-specific experience signals, and real-time trip updates so owners could feel connected without hovering.
Key design decision
Two distinct roles - plant owner and plant sitter - with entirely separate flows, dashboards, and mental models. Rather than collapsing them into one generic experience, I kept the surfaces purpose-built for each user type.
A two-sided experience built around care, not just convenience
The app supports two distinct roles, each with purpose-built flows. Owners can add their plants with detailed, per-plant care notes (light, water, soil, quirks), browse local sitters by availability and plant experience, book and confirm trips, and follow along with photo check-ins. Sitters get a clean dashboard, availability controls, step-by-step active-sit instructions, and earnings tracking.
The handoff experience was especially important to get right. Rather than a generic "start sitting" button, sitters receive a tailored walkthrough of each plant in the booking - so there's no ambiguity about what needs to happen, and owners can feel confident they've set their sitter up to succeed.
Interactive prototype - explore the full experience
On roots, growth, and building things that matter
This project was a chance to learn how to vibecode - to see what happens when you pair a designer's instincts with AI as a creative collaborator. I used Claude Code to scaffold the React Native structure, Claude and ChatGPT to pressure-test flows and generate copy, and Figma to stay grounded in what the thing should actually feel like. Watching different tools click into place for different parts of the process was genuinely exciting.
More than the tech, this was a passion project that had been sitting in my notes for a long time. I finally brought it to life - and that part felt really good. Building 0-to-1 solo is humbling: every tradeoff is yours to own, and scope creep is real when you're the designer, the PM, and the engineer. I designed flows I never built, and built flows I'd redesign with fresh eyes.
If I picked this back up, I'd invest more in the trust layer: sitter reviews, plant-specific experience badges, maybe a short video intro from each sitter. The marketplace only works if owners feel safe enough to actually book. That's the hardest design problem here - and I only scratched the surface. But for now, I'm happy to let this one grow at its own pace. 🌿