Mobile · Engineering case study
Plant Health Advisor
Client: Hydro Farm Technologies
An Android garden app that detects plant deficiencies from photos and recommends products to treat them.
The problem & context
A freelance garden-care app, published on Google Play as Genie AI, where users photograph their plants to diagnose deficiencies across 10 supported plant species.
Architecture
Plant Health Advisor is an Android app built in React Native. Gardeners upload a photo of a plant, and a TensorFlow CNN, trained offline on images of 10 plant species and exported to TensorFlow Lite, runs on the device to detect the plant’s deficiency. The app then sends the detected deficiency to a Python API, which looks up recommended products in its database and returns them to the results screen.
Illustrative system view, not a production screenshot.
Implementation & my contribution
- Built the Android app in React Native, with photo upload and results screens.
- Trained a TensorFlow CNN on images of 10 plant species and ran it on-device with TensorFlow Lite.
- Developed a Python API that maps detected deficiencies to recommended products.
Outcome
Delivered on-device plant deficiency detection with product recommendations in a single Android app.