CuratePlate

Inspiration

As broke college students who love food, we often struggle to pick affordable, filling, and (relatively) healthy options when we're going out to eat. We wanted a solution to this problem, a way for us to see how "efficient" a meal is, how much we're getting from each dollar we spend.

What it does

Our app is personalized to each user. A user will input their height, weight, age, and biological sex, which we'll use to estimate their daily nutritional needs. The user can also update any dietary restrictions they may have. The user can use their camera to take a picture of a restaurant's menu. CuratePlate will then analyze all the options on the menu and create a customized score for it based off each individual user's information. The user can select any option on the menu they want to learn more about, and the app will show them all the necessary information. There is a Meal Value score, which is a value out of 100 that represents how economically efficient the option is for its price. It is calculated based on the nutritional value of the option, which includes the micronutrients and macronutrients in the meal, and the score is lowered if the option is too high in sugar or unhealthy fats. The Meal Value score compares the nutritional value to the price of the option, and the higher the score is, the more bang for your buck it is. Additionally, the user can also see the expected general nutritional information of the option. If the user wants to learn more about a specific nutrient, they can ask Tete, our bunny mascot who's always willing to help!

How we built it

We used the Gemini API and the Nutritionx API to estimate general nutritional value for each option the user selects. We also used the Eleven Labs API to create the voice assistant of Tete. We coded in python for all the data crunching and back end, and we used Swift for all the front end stuff.

Challenges we ran into

We ran into some struggles implementing the Gemini API and the Nutritionx API to get the nutritional information about different foods. We were also new to Swift and making apps in general, so we had to iron out a lot of small hitches.

Accomplishments that we're proud of

We are proud that within the time constraints of this hackathon we were able to create a working, user friendly app that implemented computer vision. We had never made an app before, so we were proud to create an app with a pretty and easy to use interface. We were also new to implementing APIs, and while there were a few hiccups, we're proud that we were eventually able to figure it out.

What we learned

We learned a lot about designing and implementing an app, a lot about how to use computer vision, a lot about graphic design and UI design, and how to use and implement different APIs.

What's next for CuratePlate

Due to the time constraints, the scope of this project isn't as deep as we would have preferred. In the future we'd like to implement features like allowing the user to choose certain nutrients for the app to focus on. For example, someone who is trying to increase their protein intake can tell the app that they want protein to be a higher focus, so CuratePlate will adjust the rankings to give more weight to protein. We also want to allow users to choose from a wider range of dietary restrictions. Our current selection unfortunately does not include a lot of specific dietary restrictions. We hope in the future to offer a wider variety of options and allow the user to create custom restrictions.

Built With

• elevenlabs-api
• gemini-api
• nutritionx-api
• python