Virgo Stevanus
Back to projects

WellFish

Backend API & ML inference service for an AI-powered fish species identification platform.

Laravel PHP MySQL Laravel Sanctum Node.js Hapi.js TensorFlow.js GCP Docker

Preview

About this project

Capstone team project. WellFish is a mobile app that identifies 30+ fish species from photos using ML, so fishermen and traders can check fish quality and safety. I handled all the backend work.

The main API is built with Laravel 11. It connects the Android app to the ML service, handles auth with Sanctum, manages fish species data in MySQL, and stores images in Google Cloud Storage. Around 10+ endpoints covering classification, history, auth, and profiles.

I also built the ML inference service separately with Node.js and Hapi.js. It loads the TensorFlow.js model from Cloud Storage on startup and runs predictions when the API sends it images. On top of that, I containerized everything with Docker and set up the GCP architecture for deployment.

Role: Backend & ML Infrastructure Developer | Team project