Inspiration
Eating out can be convenient, but it’s often expensive and not always the healthiest option. We wanted to create a tool that allows users to compare the cost and nutritional value of their favorite restaurant dishes with homemade alternatives. This way, people can make more informed choices about where and how they eat.
What it does
BiteSMART lets users input a restaurant meal and see a detailed comparison with a homemade recipe for the same dish. It compares the cost and nutritional information (such as calories, protein, and fats) of both options. If users are interested in making the dish at home, BiteSMART also provides a recipe to help them get started.
How we built it
We built BiteSMART using a web-based stack, with HTML and CSS on the frontend for a user-friendly interface. The backend is built via Python (Flask) and uses API calls to approximate nutrition data and recipe details.
Challenges we ran into
One major challenge we encountered was difficulty in connecting to a reliable database for up-to-date nutrition and pricing information as a catch-all for meals of all restaurants. To work around this, we utilized OpenAI's API to generate estimated nutritional and cost data based on typical ingredients and portion sizes. While not a perfect substitute for a dedicated database, this approach allowed us to keep our development on track and focus on building the core functionality.
Accomplishments that we're proud of
We’re proud of creating a tool that not only provides useful information but also encourages healthier and more budget-conscious decisions. The recipe integration adds value for users interested in cooking at home. We’re also pleased with the intuitive design, which makes it easy for users to get insights in a timely manner.
What we learned
Working on BiteSMART has deepened our understanding of full-stack development and data-driven applications. We also learned how to structure and optimize the website to create a seamless user experience. Additionally, we explored the process of fine-tuning large language models (LLMs) to better align with our application's specific needs, enabling us to generate more accurate and context-relevant data when external databases weren’t accessible.
What's next for BiteSMART
In the future, we’d like to improve the overall UI/UX design and expand BiteSMART’s database to cover more restaurants and dishes, making it relevant for a wider audience and more accurate in terms of nutrition details. We also plan to integrate more recipe customization options, allowing users to tailor recipes to their preferences. We would also like to incorporate BiteSMART as a browser extension so that users are not required to search for and open the website each time to compare foods.
Log in or sign up for Devpost to join the conversation.