Plant Identifier App
An interactive web application that uses machine learning to identify plants from an uploaded image.
The project in context
Plant Identifier App lets people capture or upload a photo and receive an AI-powered plant identification. This personal project explores how to make an image-based AI feature feel clear and approachable.
Plant identification tools can make a simple question feel complicated. The experience needed to make the first step obvious while still setting expectations around image quality and AI-generated results.
My role & stack
Full-stack development and AI integration
NuxtTypeScriptGemini API
What I built
- Designed a focused upload flow around camera capture and image selection.
- Connected the interface to the Gemini API for image-based identification.
- Implemented responsive states for empty, loading, successful, and failed identification attempts.
Implementation notes
Technical approach
- A single primary action keeps the camera and upload flow easy to discover.
- Typed request and response boundaries keep the AI integration predictable.
- Guidance beneath the controls helps people submit a clear photo before they start.
Takeaways
- AI features feel more useful when the interface makes the next human action unmistakable.
- Small pieces of upfront guidance can improve the quality of an image-based result.