WildGuard
AI-powered wildlife rescue platform connecting people with the right NGOs instantly.
What it does
A platform that helps people who find injured or distressed wildlife quickly locate and contact the most relevant NGOs, shelters, and volunteers in their area. Users can photograph the animal and the platform identifies the species and recommends suitable organizations.
Why I built it
Wildlife rescue in India is fragmented — most people don't know who to call or which shelter handles which species. The gap between finding an injured animal and reaching the right help costs lives. WildGuard closes that gap with AI and a searchable NGO directory.
How it works
Users authenticate via Firebase, upload a photo, and the backend runs it through Google Cloud Vision for object detection and Google Gemini API for multimodal species identification and context understanding. Results are matched against a PostgreSQL database of registered NGOs and volunteers, filtered by location via Google Maps. NGOs and volunteers can register themselves through the platform.
Tech Stack
Frontend — React, Vite, TypeScript, Tailwind CSS, ShadCN, Firebase Auth. Backend — Express.js, TypeScript, Prisma ORM, PostgreSQL (Supabase dev / GCP prod), Google Gemini API, Google Cloud Vision, Firebase Admin SDK, Google Cloud Storage, Docker.