← Back to Projects

An AI-assisted mold identification and monitoring platform that helps farmers detect and monitor mold growth through image analysis and metadata.

Highlights

• Orchestrated a multi-platform REST API on Firebase Functions and AWS Lambda, with CI/CD via GitHub Actions, connecting the backend and ML service to the website and mobile application.
• Finetuned and built a multimodal CNN-ANN mold image classification model trained on 571 labeled microscopic images across 6 genus categories, achieving 93% overall accuracy on the CNN layer.
• Implemented local caching (Redis/SWR/Dio) across the API, Next.js web, and Flutter mobile clients.

View on GitHub