About the Project

Next Earth was inspired by the urgent need to make climate action accessible and equitable - especially for youth in regions most affected by floods, droughts, and heat risks. We wanted to answer one simple question:

“How can anyone, anywhere, turn their skills into meaningful climate impact?”

That question led us to build an AI-powered climate-action navigator that connects people’s skills and locations to real roles and learning opportunities, using open climate data such as ASDI and NOAA.

✨ What We Learned

Throughout the project, we learned how to:

Integrate AI reasoning into a real-world web app using OpenAI’s API.

Combine structured data (climate risk scores, skills) with unstructured user input to deliver meaningful recommendations.

Design for clarity and empathy - using soft UI, transparent explanations (“Why this fits”), and accessibility-friendly layouts.

Translate UN Sustainable Development Goals (SDG 10 & 13) into a tangible, user-driven solution.

🧩 How We Built It

We built Next Earth in three phases:

Frontend & UX: Using Next.js with TypeScript, Tailwind, and CSS Modules for a modern, responsive, and glassmorphic design.

Matching Engine: Custom logic that matches user skills and chosen “path” (health, community, green workforce) with curated roles and risk-weighted data.

AI Integration: The /api/adviser route connects to OpenAI’s GPT-4o-mini, giving contextual advice and micro-learning recommendations based on the user’s country and skills.

⚙️ Challenges We Faced

Time & integration pressure: Merging multiple screens (Profile + Suggestions) into a single, reactive “Matches” page under tight deadlines.

Model reliability: Handling fallbacks and response limits when OpenAI API throttled or failed.

Data realism: Simulating ASDI/NOAA-style datasets while keeping results grounded and explainable.

Team collaboration: Managing fast commits and avoiding overwrites during Git merges.

Built With

Share this project:

Updates