Mangrove Guardian AI — Protecting Coastal Ecosystems with AI

EL MOUATAZ BENMANSSOUR — Sun, 19 Apr 2026 16:43:23 +0000

This is a submission for Weekend Challenge: Earth Day Edition

The Problem

Mangrove ecosystems serve as one of the most powerful defenses to counter both climate change and coastal erosion.
Unfortunately, monitoring mangrove damage and degradation is problematic in that reports are often scattered across local authorities, NGOs, and communities.

Efforts to restore mangrove ecosystems face some challenges:

real-time field reporting is limited
there is little coordination amongst communities and organizations
evaluating environmental damage takes a long time.

What I Built

Mangrove Guardian AI is a community-driven platform that helps people protect mangrove ecosystems using reporting, geospatial data, and AI-assisted analysis. The platform incorporates crowd-reported data through manual photographic evidence, geographical location data, and an artificial intelligence (AI) analytical component.

The application is designed to transform environmental reporting into actionable items rather than simply providing information. The system will convert every submitted report (including photographic evidence) into an organized collection of signals that organizations may utilize to accomplish real restoration opportunities in mangrove areas.

The platform provides the following functionality:

Provides individuals an opportunity to provide reports on damage to mangroves and upload photographic evidence and geolocation-based information.
Utilizes an artificial intelligence-based analytical module to evaluate incoming reports and provide corresponding health and risk scoring of each mangrove area based on the evidence that has been submitted.
Provides access to unique dashboards (by avatar) for community and organizational users, based on the role(s) that they play within the platform.
Facilitates tracking of restoration and field-based activities over a designated time period.
Provides reporting features that can be exported in order to allow for improved coordination and accountability across restoration projects in the mangrove areas.

Demo

Live Demo:

mangrove-guardian-ai-lcmu.vercel.app

Video Walkthrough :

Code

moatazbenma / Mangrove-Guardian-AI

🌳 Mangrove Guardian AI

A comprehensive web platform for reporting, monitoring, and restoring mangrove ecosystems using AI-powered analysis and community engagement.

Status: Production-Ready | Version: 1.0.0

📋 Project Overview

Mangrove Guardian AI enables communities and organizations to:

Report mangrove damage with geolocation and photographic evidence
Analyze damage using AI inference to assess health scores and risk levels
Track restoration projects and measure environmental impact
Collaborate across community and organizational user roles

🎥 Demo

Project demo video: https://youtu.be/aZBV_R53PRY

🏗️ Tech Stack

Frontend

React 19 + TypeScript
Vite for fast development & bundling
Tailwind CSS v4 for responsive UI with custom eco-theme
React Router for navigation
React-Leaflet for interactive maps
Axios with JWT authentication

Backend

Django 4.2.16 LTS + Django REST Framework
Python 3.11
Gunicorn WSGI server (4 workers)
SimpleJWT for token-based authentication
PostgreSQL 15 (production) / SQLite (development)

Services & Infrastructure

Redis 7 for caching and rate limiting
Celery 5.3.4 for…

View on GitHub

System Architecture

How I Built It

I built the project as a containerized, service-oriented web app with clear separation between UI, API, async processing, and infrastructure.

Frontend

React + TypeScript + Vite
Tailwind CSS for responsive styling
React Router for navigation
React Leaflet for map-based context
Axios with JWT handling for authenticated API calls

Backend

Django + Django REST Framework
Custom user roles and JWT authentication
Modular apps for users, reports, analysis, and restoration
Pagination-friendly API responses for scalable listing views

AI + Async Processing

Celery workers process analysis tasks outside request/response flow
Redis used as broker/cache to keep operations fast and resilient
Analysis outputs include health scoring and risk indicators for triage

Data + Storage

PostgreSQL in production-style setup
Cloudinary for image storage and delivery
OpenPyXL export support for organization-level reporting workflows

DevOps + Deployment Readiness

Docker Compose setup with multiple services:

Backend API
Frontend
PostgreSQL
Redis
Celery worker
Celery beat

Prize Categories

I am submitting this project for:

Best Use of GitHub Copilot GitHub Copilot helped accelerate implementation across frontend components, API integration, and backend endpoint scaffolding while keeping iteration speed high during weekend build constraints.

This project was built as part of the Earth Day Weekend Challenge to explore how technology can help communities protect critical coastal ecosystems.

DEV Community: EL MOUATAZ BENMANSSOUR