DEV Community

Cover image for GAIA: The Interactive 3D Earth Assistant | Earth Day 2026
Gbolahan Olanrewaju
Gbolahan Olanrewaju

Posted on

GAIA: The Interactive 3D Earth Assistant | Earth Day 2026

DEV Weekend Challenge: Earth Day

This is a submission for Weekend Challenge: Earth Day Edition

What I Built

Over 1.3 billion tons of food is wasted globally every year, contributing massively to greenhouse gas emissions. Carbon footprint calculators are great, but for this Earth Day challenge, I wanted to build a literal connection to the planet that solves a specific, actionable problem: food preservation and waste reduction.

I built GAIA (Global Agricultural Intelligence Assistant).

GAIA is an interactive, sentient 3D Earth that acts as an intelligent food lifecycle and safety manager. Bringing my background in Food Science and Technology into the software space, I designed GAIA to move beyond generic sustainability tips. Instead, GAIA provides scientifically grounded advice on extending vegetable shelf life, the safety and functional roles of fermenting condiments, and upcycling surplus food before it spoils.

Demo

🌍 Live Application: https://gaia-earth-buddy.vercel.app/

(Note: Try clicking the quick-action runes on the right, or ask GAIA directly how to safely upcycle overripe fruit!)

See GAIA in action:

Code

💻 GitHub Repository: gbollybambam/gaia-earth-buddy

How I Built It

I wanted an immersive, borderless, and premium UI. I built the frontend using Next.js and react-globe.gl to render a highly performant, interactive 3D WebGL Earth. The glassmorphic terminal UI was built from scratch using custom CSS to float over the planet.

For the brain, I utilized the cutting-edge Gemini 2.5 Flash model via the @google/generative-ai SDK.

Technical Highlights:

  • Strict JSON Formatting: I engineered the system prompt and responseMimeType to force Gemini to return strict JSON containing both a text response and a specific mood state.
  • Stateful UI: The AI's returned mood dynamically triggers CSS changes on the frontend, making the globe react emotionally to the data (e.g., expressing joy when discussing upcycling, or concern when analyzing global food waste).
  • High-Speed Architecture: By utilizing the 2.5 Flash model and edge-ready API routes, the data streams are incredibly fast, creating a seamless, conversational feel.

Prize Categories

  • Best Use of Google Gemini

Top comments (0)