Inspiration
We aim to build something meaningful for the “Data Science for Social Good” theme. Everyone has dinner, but very few people know how much CO2 their meal produces. Our inspiration was to make carbon emission awareness easy, personal and recognizable to all people.
What it does
Our tool lets users:
- build a dinner plate by selecting foods and portion sizes
- instantly see the CO2 emissions for each ingredient and the entire meal
- Discover the alternatives that lessen CO2 emissions
- understand how everyday food choices affect us
- compare items visually with descriptive and real world breakdowns
How we built it
Data:
- we found a data set of foods with affects of CO2, land and water
- we cleaned the data and computed the dataset per portion and other useful features
Frontend: React + Typescript + Vite + HTML + TailwindCSS TailwindCSS was used for fast styling
Backend: Our backend is built with FastAPI and Python3. Gemini integration. our programs provide:
- ingredient data endpoint
- plate submission endpoint
- a full environmental impact computation
- cross origin support for frontend integration using FastAPI we check type safety and apply fast responses
Core logic we created a custom script to Load the dataset, clean it, normalize, and compute metrics. we then added default portion sizes, categorized items and exported a clean json file used by the backend.
Challenges we ran into
- Finding reliable CO2 Data for food items due to datasets variations available
- building a usable interface in a short time
- creating calculations that stayed accurate
- balancing speed and detail so users don't feel overwhelmed
Accomplishments that we're proud of
- we turned a complicated topic into something simple and interactive
- built a working react app from scratch using Vite + TypeScript
- made climate action feel accessible, not political or intimidating
What we learned
- how to transform raw environmental data into a user friendly app
- how different foods drastically differ emissions
- how to structure data pipeline from cleaning to feature engineering (UI)
- how small design choices dramatically affect clarity and usability
What's next for Dinner and CO2 Emissions
- Expand the food database
- add cultural foods
- introduce meal suggestions with CO2 alternatives
- Build mobile version
Log in or sign up for Devpost to join the conversation.