Fit Compass

Guides you in all directions of fitness, from calories to form.

Comment

Inspiration

Our team is super into fitness and working out, and we all play sports, so we thought it would be a great idea to make a fitness app. We know getting started with fitness can be tough, so we wanted to make something that’s easy to use and super helpful. Our app isn’t just about the gym—it calculates calories, helps with form to avoid injuries, provides healthy recipes, and offers tips to stay on track. We’re passionate about encouraging people to build good habits and feel confident in their fitness and nutrition journeys, regardless of their starting point.

What it does

FitCompass is designed to guide users in all aspects of a healthy lifestyle. It calculates calories to help users stay on top of their nutrition goals, provides personalized workout plans, and includes a bot that provides live tips for proper form to complete exercises safely. The app also offers a specialized chatbot to make meal planning easier. With a focus on promoting fitness, nutrition, and overall wellness, the app is perfect for anyone looking to improve their health, from beginners to experienced athletes.

How we built it

We created a linear regression model that took into account age, gender, current weight, amount of weeks they want to take, how active they are, and the weight they want to achieve. Using this, we predicted the amount of daily calories to achieve weight. After the prediction, we offered a macronutrient breakdown from the calories as a suggestion. We also created a health and fitness chatbot that specializes in creating customizable recipes. We implemented this using the Anthropic Claude-3 Sonnet v1..0 API in AWS Bedrock. Lastly, we created a live squat exercise form corrector and repetition counter using OpenCV.

Challenges we ran into

We ran into a bunch of challenges while building FitCompass. Finding a dataset for our predictive model was harder than we expected, and even when we got it, we struggled with the model's accuracy. We started using Django but ran into problems, so we had to transfer all our code to Flask, which took a lot of time. On top of that, the front-end work was tough—we struggled with HTML, CSS, and JavaScript, especially figuring out animations and where things should appear on the page. It was a lot to handle, but we learned so much along the way!

Accomplishments that we're proud of

We are extremely proud of the various models we created and how the app offers different applications of AI for the specific focus we chose. We are also proud of the implementation of all our models in the front end and how we were able to bring it all together.

What's next for Fit Compass

We want to make it even better by improving the accuracy of our predictive model with more data, so the app can give even smarter recommendations. We’re also planning to create custom workout plans based on each user’s goals and add more exercises that include live-form assistance. Plus, we’ll connect the live form help directly to the custom plans, so it all works seamlessly. To keep improving the app, we’re thinking about adding a subscription option for extra features and support!

Built With

Share this project:

Updates