The problem CarbonTrace solves

CarbonTrace helps individuals and communities reduce their carbon footprints by providing a comprehensive suite of tools that integrate seamlessly into their daily lives.

Eco-Friendly Decision Making: With features like OCR Integration and Audio Processing, users can easily extract actionable insights from documents and audio recordings related to sustainability, enabling them to make informed decisions about their carbon impact.

Personalized Carbon Footprint Tracking: The Carbon Footprint Calculator helps users track and understand the environmental impact of their daily activities, such as diet, travel, and shopping. This personalized feedback encourages users to adopt more sustainable habits by offering insights and alternative choices.

Carpooling and Travel Coordination: The Carpool Map allows users to coordinate rides with others in their local community, minimizing the number of cars on the road and reducing emissions.

Sustainable Shopping: The Green Product Recommendations feature makes it easier for users to find eco-friendly alternatives when shopping online by redirecting them to products that align with their sustainability goals.

Gamified Engagement: Through the Leaderboard and Rewards System, users are motivated to lower their carbon footprints in a fun, competitive way. This feature adds an incentive for users to continuously engage with the app and make eco-conscious decisions.

Event Participation and Carbon Offsetting: Event Registration and Carbon Offsetting options give users tangible ways to actively contribute to sustainability. Users can participate in carbon offset programs and attend events focused on reducing their environmental impact.

Challenges we ran into

Browser Extension Compatibility:
Developing a browser extension that integrates seamlessly with Amazon and redirects users to eco-friendly alternatives proved tricky due to the complexities of parsing Amazon’s constantly changing product listings. We overcame this by implementing a dynamic scraper and API-based solution that ensured the extension could adapt to frequent changes in product data, delivering up-to-date recommendations.

Lack of GPU Computing Power:
We faced significant challenges related to insufficient GPU computing power, especially when running resource-intensive models like Whisper for audio transcription and LangChain for semantic search. The lack of GPU support resulted in slower processing times and limitations in scaling the system for real-time, large-scale data processing. To overcome this, we relied on cloud-based GPU services during critical processing phases, allowing us to offload intensive tasks while optimizing local resources for less demanding operations.