GreenPrint

Inspiration

We wanted to bridge the gap between personal finance and environmental awareness. Most people know how much they spend--but not the carbon impact of their lifestyle. By reframing everyday transactions as carbon spending, we empower individuals to make more sustainable choices through data-driven insights.

What it does

GreenPrint connects to your bank (via Capital One’s Nessie API) and automatically estimates the CO_{2}e emissions behind each transaction. It visualizes your “carbon budget” across spending categories, tracks monthly trends, and generates personalized recommendations--like switching to public transit or supporting low-emission retailers.

Each recommendation is powered by Echo (Merit Systems) -- a unified API router for AI models -- which we use to route and manage calls to the OpenAI API. Echo enables GreenPrint to generate smart, AI-driven sustainability insights and personalized actions based on each user’s spending patterns, without exposing model API keys or managing multiple LLM providers directly.

How we built it

We built a full-stack TypeScript monorepo using Next.js 14 and Express + tRPC.

Database: PostgreSQL with Prisma ORM
Auth: JWT-based authentication
Data Analysis: Local emission factors (EPA, DEFRA)
AI Insights: Integrated Echo (Merit Systems) to securely route model requests to OpenAI API for generating sustainability recommendations
Visualization: Recharts + Tailwind + shadcn/ui
Infrastructure: Docker-ready and deployed on Vercel
CI/CD: GitHub Actions + pnpm + Turborepo for automated builds and deployment

Challenges we ran into

Mapping financial transaction data to accurate carbon categories
Designing CO2e estimation models that remain transparent and traceable
Integrating Echo for AI-driven recommendations while maintaining type safety
Managing client-side authentication and routing with Next.js App Router
Balancing production-level architecture with the limited time of a hackathon

Accomplishments that we're proud of

Built a fully functional FinTech sustainability app with real-time carbon footprint estimation
Seamlessly integrated Echo + OpenAI for intelligent, personalized recommendations
Achieved full-stack type safety with tRPC and Prisma
Deployed a production-grade application on Vercel with working CI/CD
Created a visually engaging and transparent dashboard that demystifies carbon data

What we learned

We learned how financial APIs and sustainability data can intersect to drive climate awareness. Integrating Echo taught us how to abstract LLM calls securely and efficiently, while still delivering intelligent, meaningful insights. We also deepened our understanding of full-stack TypeScript, serverless deployment, and sustainable software design.

What's next for GreenPrint

Add real Capital One and Plaid API integrations for live financial data
Expand Echo usage to support multi-model AI pipelines (e.g., Gemini + Claude + GPT)
Introduce gamified carbon goals, community challenges, and social leaderboards
Add carbon offset tracking and credit purchasing directly from the dashboard
Continue optimizing for scalability with managed Postgres (Neon or Supabase)

💚 GreenPrint -- turning your financial data into actionable climate impact, powered by Echo and OpenAI AI.