With two friends, we worked on a project for 24 hours at the Berkeley AI Hackathon. I programmed the vector embedding system which used OpenAI's embedding model to calculate the similarity of products. This allowed us to find similar products with a lower carbon footprint while avoiding hallucinations from the AI. We won the climate track of the hackathon and also received $25k as an investment from Berkeley Skydeck, with whom we are working with to further our product.