AI · Full Stack · Agriculture

TomatoCare

AI-powered plant screening,
thoughtfully built.

An AI-powered full-stack platform that helps users screen tomato leaf diseases through image analysis, secure workflows and clear reporting.

Live demo GitHub

AngularExpressFastAPIPyTorch

Overview

Making AI-assisted plant
screening more accessible.

TomatoCare is a full-stack web application that allows users to upload tomato leaf images and receive AI-assisted disease screening results through a clear and accessible workflow.

The platform combines image analysis, secure user accounts, analysis history and downloadable reports within one integrated experience.

Tech stack

Frontend

Angular

Backend

Node.jsExpressJWT

FastAPIPyTorch

Data & delivery

MongoDBDocker

Key features

AI Disease Screening

Upload tomato leaf images and receive AI-assisted disease predictions with confidence information and alternative predictions.

Uncertainty Awareness

Ambiguous or low-confidence predictions can be identified as uncertain and flagged for additional review instead of being presented as certain results.

Dashboard & Analysis History

Authenticated users can review previous analyses, track results and manage their personal report history.

Downloadable PDF Reports

Generate structured reports containing the analyzed image, prediction information and relevant screening details.

Architecture

Built as independent services with clear responsibilities.

TomatoCare separates the user interface, application API and AI inference responsibilities into independent services.

The Angular frontend communicates with the Express API, which manages authentication, report persistence and communication with the dedicated FastAPI/PyTorch inference service.

AngularFrontend

Node.js / ExpressAuthentication · business logic

Persistence

MongoDBUsers · reports

AI request

FastAPI AIPyTorch · ResNet50

Application workflow