[amgad]

AgriWise

Top 50, Google APAC Solution Challenge 2025

AI agriculture platform that helps Indonesian farmers detect plant disease, assess seed quality and predict pest growth. Dockerized Node.js backend on Google Cloud Run with Firestore, Cloud Storage, Firebase Auth and Gemini, serving a Flutter app.

ROLE
"Backend lead"
PERIOD
"Apr – May 2025"
CONTEXT
"Google APAC Solution Challenge, GDGoC UII"
TEAM
"4 people"
STACK
"Node.js, Cloud Run, Docker, Firestore, Gemini"
UPDATED
"Jun 2025"
github.com/amga-d/AiMpostor

Overview

AgriWise helps Indonesian farmers catch problems early. They photograph a plant and get a diagnosis with treatment advice, check seed quality, get fertilizer recipes and see pest-growth forecasts.

I built the backend for a four-person team: a Node.js API that takes image uploads from the Flutter app, sends them to Gemini for analysis and stores the results in Firestore. The project was selected for the top 50 in the Asia-Pacific region.

What I built

  • REST API in Node.js, with image uploads handled by Multer.
  • Gemini integration for disease detection, seed assessment, fertilizer recipes and chat.
  • Firestore data model, Cloud Storage for images and Firebase Auth for accounts.
  • Dockerfile and deployment to Google Cloud Run, with environment configuration for production.

How it fits together

    • Flutter appFirebase Auth
  1. HTTPS
    • Node.js APIDocker on Cloud Run
  2. Google APIs
    • Geminiimage analysis
    • Firestoreapp data
    • Cloud Storageplant photos

In the repository

$ git clone https://github.com/amga-d/AiMpostor.git

tree -L 1

.
├── backend/
├── frontend/
├── .gitignore
├── LICENSE
└── README.md

git log --author=amga-d --oneline -5

  1. 6037b7cfeat: add secrets directory to .gitignore
  2. f41d1a3remove redundant files
  3. a75356adocs: update backend documentation
  4. 73029f3docs: Update frontend/Agriwise/README.md with overview, features, tech stack, and setup instructions
  5. dc8da7erefactor: Correct endpoints paths adding leading slashes
  • Dart 68.7%
  • JavaScript 14.4%
  • C++ 8.4%
  • CMake 6.7%
  • Other 1.7%