EcoMind ai

Inspiration

Most sustainability apps are either hard to use or do not show real, measurable impact. We wanted to build a simple product that feels practical for everyday users, while still using a real cloud data platform for analytics.

What it does

EcoMind AI helps users:

Estimate carbon footprint from everyday habits (travel, electricity, food, waste)
Get personalized sustainability recommendations
Classify waste with a simple upload flow (mock classifier)
Track impact through a climate dashboard with analytics
Stay engaged using points, badges, and a leaderboard

Databricks-powered functionality in this version:

Stores hackathon sample event data in a Databricks table
Runs SQL aggregation on Databricks SQL Warehouse
Fetches live total points through a server-side API route
Displays live Databricks values directly in the dashboard with source labeling

How we built it

We built EcoMind AI using:

Next.js (App Router) + TypeScript
Tailwind CSS for a clean, minimal UI
Framer Motion for lightweight transitions
Recharts for simple analytics visualizations
Databricks Free Edition (SQL Warehouse + SQL Statements API)

Architecture highlights:

Modular reusable components
Local logic for analysis and recommendations for fast prototyping
Clear separation of concerns across app pages, components, and lib utilities

How Databricks is used end-to-end:

We created a schema and table for eco events in Databricks
We inserted sample user points rows for the demo
We added a server route in the app to execute SQL via Databricks SQL Statements API
The dashboard requests this route and updates the Points card with live Databricks totals
We added a visual indicator in UI to show the data source is Databricks

Databricks-Specific Details

Where Databricks is used:

Data storage: hackathon.eco_events table in Databricks
Analytics: SQL aggregation (total eco points) executed on Databricks SQL Warehouse
App integration: server-side Next.js API route calls Databricks SQL Statements API
UI integration: dashboard Points card and status indicator show live Databricks-backed values

What worked great with Databricks:

Fast setup on Free Edition for a hackathon timeline
Simple SQL workflow for creating tables and running aggregations
SQL Statements API made backend integration straightforward
Good fit for showing real cloud-backed analytics instead of only static mock data

What was frustrating / challenging:

Initial environment variable setup and token handling required careful backend-only configuration
Minor trial-and-error around endpoint wiring and local dev verification
Free-tier/runtime constraints mean we optimized for a small, focused live metric

Challenges we ran into

Connecting Databricks securely from a Next.js app without exposing tokens on the client
Managing environment variables correctly so local development and API routes could read credentials
Handling package/runtime compatibility issues while keeping the dashboard stable
Balancing fast hackathon delivery with clear architecture and clean code

Accomplishments that we're proud of

Delivered a full end-to-end working product with multiple climate features
Kept architecture scalable and clean for future upgrades
Successfully integrated Databricks Free Edition into the live dashboard flow
Added visible proof of live cloud data usage in the product UI
Completed integration with minimal complexity suitable for rapid hackathon execution

What we learned

Strong modular architecture speeds up feature delivery and debugging
Simple rule-based AI logic can still provide clear user value in early versions
Databricks SQL APIs are quick to integrate for real-time analytics use cases
Secure server-side integration patterns are critical, even for demos
Build and environment reliability are as important as feature implementation

What's next for EcoMind ai

Expand Databricks usage from one live metric to full dashboard pipelines
Add scheduled ETL jobs on Databricks for daily/weekly sustainability summaries
Introduce user-level history tables and cohort analysis in Databricks
Integrate Databricks notebooks and model serving for richer recommendations
Add authenticated user accounts and long-term progress tracking
Improve waste classification with a real computer vision model and feedback loop
Launch team-based challenges and organization-level climate goals