Climate Conscience

Our Landing Page

Inspiration

In the last week, I made a realization I once thought to be impossible, Bay travel is completely possible using only public transportations. Just in the last 9 days I have spent over 15 hours on bart and other forms public transit and successfully used it as my only source of transportation. When it comes to doing climate good, taking public transit is one of the best ways you can reduce your emissions. However, there is much more to climate good and that is where Climate Conscience comes into play....

What it does

Climate Conscience is designed to both educate on and gamify doing climate good. The goal of Climate Conscience is to create a powerful knowledge base using public/private information on causes of climate change in the Bay Area, laws that enforce and/or encourage behavior that benefits the surrounding climate. We engage with our user by prompting them to take our climate responsibility survey and then we query our knowledge base to provide the user with feedback on how they can improve their score. In order to encourage growth and build user engagement we have curated a list of challenges our users can take on. Each challenges tracks users performance and aggregates the user's performance as well as the performance of our global user base.

How we built it

We utilized the next.js framework, aws bedrock, aws dynamoDB, Claude, bcrypt, jwt, charkaUI, and MongoDb to build Climate Conscience.

Challenges we ran into

It was incredibly challenging to complete Climate Conscience under the time limit with only two people, especially when we were encountered a new error that one of us had never seen before and time spent pair programming or debugging meant nobody was coding. Both the AWS Bedrock Javascript SDK and bcript had critical errors in the SDK itself which took hours to debug and find work arounds.

Accomplishments that we're proud of

We were committed to coding more than ever before, this marked our first hackathon where we actually stayed at the venue overnight and I ended up having a dream during a nap where I solved the error I was debugging.... it was at that level. Speaking of debugging, we are super proud of all the debugging that we did do, adapting where needed and successfully implementing AI in Javascript where everything AI is done in python these days.

What we learned

We learned how to call a LLM in Typescript and learned our mistakes in procrastination on implementing the projects bottleneck as once we got to it and encountered issues all of the sponsors and mentors had gone home for the night. It also might not hurt next time to check our package versions in advance so we know if we can trust the current version or not.