Inspiration Climate change is one of the biggest challenges of our time — yet most people don’t understand how it affects them. We were inspired to build something that could bridge the gap between complex climate data and everyday understanding. What if AI could help people explore real-world climate trends in plain language and see it unfold on a global map?
🛠️ What We Built Climate Insight Explorer is a full-stack AI-powered web app that answers climate questions using real data and shows trends on an interactive heatmap. Users can ask natural language queries like "What happened during the California drought in 2022?" and instantly get concise, AI-generated answers — backed by data from NASA’s GISTEMP dataset.
It features:
🌡️ Gemini/Gemma-generated summaries of climate events
🗺️ Google Maps integration with heatmap & toggleable circle/marker view
🔎 AI + real-time MongoDB queries to ground insights
🔁 Full interaction history storage + visualization
🚧 Challenges We Faced Fine-tuning prompts to keep AI answers grounded in data
Matching region names from user input with MongoDB documents
Styling a clean UI while integrating multiple JS layers
Making the map switchable (heatmap vs. markers) in real time
🙌 What We Learned How to integrate Google Maps + MongoDB geospatial data
How to serve AI queries dynamically using Gemini and Python Flask
Handling and visualizing structured + unstructured data together
Working fast under pressure while building clean and extensible code
Built With
- apis)-??-mongodb-atlas-(climate-dataset-+-vector-storage)-?-flask-(python-backend)-?-tailwind-css-+-html-(frontend-ui)-??-google-maps-javascript-api-?-dotenv
- gemini-/-gemma-ai-(text-generation)-??-google-cloud-(deployment
- json
- rest
Log in or sign up for Devpost to join the conversation.