This project experiments with generative AI to create an engaging and educational wildlife identification experience, by allowing users to roleplay as a naturalist from the 1800s. The visual identity draws inspiration from scientific illustrations of the era. I acknowledge the limitations of this technology in terms of factuality, and ground generated content on open access resources whenever possible.
🌐 Web App: edouardleurent.com/nature-go
Nature GO is a Progressive Web App that can be installed on your device for a native app-like experience:
iOS (Safari)
- Open the web app in Safari
- Tap the Share button (square with arrow)
- Scroll down and tap Add to Home Screen
- Tap Add to confirm
Android (Chrome)
- Open the web app in Chrome
- Tap the three-dot menu (⋮)
- Tap Install app or Add to Home Screen
- Tap Install to confirm
- Inspired by the Biodiversity Heritage Library and Pokémon GO
- Plant data from GBIF
- Bird data from Avibase
- Species details from Wikipedia
- Plant identification with Pl@ntNet API
- Bird identification with Gemini 3 Flash
- Illustration generation with Imagen 3
- Description and quiz generation with Gemini 3 Flash
- Backend with Django, frontend with Next.js