prathik@chicago:~$

EnsoGrow

A gardening app that diagnoses sick plants from a photo. Built with a team at WildHacks 2025.

when
Apr 2025
where
WildHacks 2025 (hackathon, team project)
role
Team member
stack
Next.js, React, Tailwind CSS, Gemini API, Computer vision, AWS
links
github · live demo
problem
Gardening apps ignore your space, light and time, so beginners give up when the advice feels generic.
built
A PWA that matches plants to your space, diagnoses a sick plant from a photo with Gemini, and sends care reminders.
result
A working prototype built over the hackathon and still live on Vercel.
Contents
The problem
What it does
Flows
Not done yet

The problem

Most gardening apps ignore local constraints like space, light and time, and they overlook organic-first care. Beginners get confusing advice, lose a crop, get little feedback, and give up.

We built EnsoGrow at WildHacks 2025 as a coach rather than a manual, so it gives short steps instead of long instructions. It is aimed at people growing food at home: urban growers, students and busy professionals.

What it does

  • Plant matchmaker: suggests plants that fit your space, your sunlight and the time you have.
  • Plant doctor: take a photo of a sick plant and get a diagnosis with organic treatment tips.
  • Care reminders, based on the growth stage, the weather and your schedule.
  • A growth tracker with a timeline and milestone badges, to keep you going.
  • A local community for things like compost pickups, seed swaps and tips.
  • A chatbot for gardening questions.

Flows

How a person moves through the app, and what each task involves

The user flow follows one person through sign-in, setup, plant picks, the dashboard and the plant doctor, with the front end and Gemini in their own lanes and the error paths marked. The task flow breaks the same journey into four tasks. Both are long, so click to zoom.

EnsoGrow user flow in three lanes: the user, the front end, and Gemini, from landing page through onboarding, recommendations, dashboard and plant doctor, with auth, camera and Gemini error paths
Fig. 1 User flow across the user, the front end and Gemini.
EnsoGrow task flow in four tasks: sign in and set up, start growing a plant, update plant health, and diagnose with the plant doctor
Fig. 2 Task flow: setting up, starting a plant, updating its health, and the plant doctor.

Not done yet

  • Testing with more growers across different climates and living spaces.
  • A better vision model, trained on more labeled plant-disease data.
  • A deeper community side, such as seed swaps, compost pickups and expert office hours.

< back to projects/