MacroSnap

Logo
Home Page
About
Options

Inspiration 🤔

In today’s fast-paced world, finding the right foods to support a healthy lifestyle can feel overwhelming, especially with the constant flood of diet trends and conflicting information. MacroSnap was created to simplify this process by offering personalized meal suggestions based on your unique needs, preferences, and available ingredients. Our app takes the guesswork out of healthy eating by providing easy-to-follow recipes and detailed macronutrients of all meals you eat. With MacroSnap, you can make informed food choices that align with your wellness goals, all while saving time and reducing food waste.

What It Does 🔍

MacroSnap simplifies healthy eating by providing personalized meal suggestions based on dietary needs, available ingredients, and health goals. Users complete a short questionnaire, which helps us provide their best recommendations. Using generative AI, we provide exceptional recipes for users to enjoy. Also, using image recognition, we provide detailed macronutrients on the foods you are eating, guiding users to their goals of healthier eating.

How We Built It 🛠

Front End: The visual aspects of MacroScap were created using Streamlit, Python, and CSS. This allowed our interface to provide functionality and aesthetics to our user-friendly website.

Back End: Our back end's structure was built upon Python and Cohere API to create personalized recipes for the user to reach their unique goals. The Cohere API processed the information given from the questionnaire and created personalized results. Also, we used Firebase to store users' information in a secure database. Lastly, we use Ollama for the image recognition part of our system.

Challenges We Ran Into ❌

Although we ran into many challenges along the way, here are some of the most notable ones:

Making the image recognition work thoroughly was a great challenge as the models were not capable of differentiating food very easily. This caused many troubles when testing the module and trying to figure out a solution. The program does the job decently, however, there is still room for improvement as results could be more accurate.
When using the generative AI in our program, it takes a very long time to run and execute effectively. Unfortunately, this issue still does remain but we have improved the issue a bit as the program does run a bit more efficiently.

Accomplishments That We're Proud Of 🎉

Here are some of the accomplishments we are most proud of:

We learned how to use generative AI to provide insight recommendations for the user in terms of their health goals. This made us familiar with the Cohere AI.
We learned Streamlit and the front end we can create using its software.
We figured out how to use image recognition to provide decently accurate information and results to users about their food intake.

What We Learned 🧠

From this experience, we learned how to make use of Streamlit to deploy websites, implement Cohere AI successfully, and make use of Image Recognition to accurately depict images such as different foods.

What's Next For MacroSnap 🚀

Implement a sophisticated calorie tracker that keeps track of all food inputted in the system, creating better recommendations based on daily feedback.
Improving loading times for the generative AI, creating more seamless and user-friendly software.
Make our image recognition more accurate to provide better results to the user about their food intake.