.../Web Application...
Crop Element
Scalable REST API backend serving ML models for crop recommendations and plant disease detection (~87% accuracy).
const project = {
title: "Crop Element",
category: "Web Application",
stack: ["Python", "Flask", "React", "REST APIs"],
}
.../overview...
Designed and implemented a scalable REST API backend serving ML models — a Naive Bayes classifier for crop recommendations and an Inception v3 transfer learning model for plant disease detection (~87% accuracy). Built API contracts with proper error handling, input validation, and response formatting consumed by the React front-end, with fallback handling for network failures and model inference timeouts.
Problem
The product needed reliable recommendation and disease-detection workflows so growers could act on data instead of manual intuition.
Role
Backend-focused full-stack engineer (University Project)
Implementation highlights
- 01Designed and implemented scalable REST API endpoints serving Naive Bayes crop recommendation and Inception v3 disease detection models.
- 02Built API contracts with proper error handling, input validation, and response formatting.
- 03Added fallback handling for network failures and model inference timeouts.
Outcomes
- ›Delivered plant disease detection at ~87% accuracy via the Inception v3 transfer learning model.
- ›Provided a single API layer powering both crop recommendation and disease analysis workflows.
- ›Enabled the React front-end to consume reliable, well-formatted ML inference results.