Rhythm AI

With Rhythm, every workout tells a story. Dive into personalized feedback sculpted by advanced AI, echoing a trainer's touch.

Comment

Inspiration

In the evolving world of AI technology, we recognized an opportunity to enhance everyday experiences. One area that felt close to home was our post-workout interactions with GPS watches. The monotony of recurring, templated comments was evident. We envisioned an AI-driven platform that echoed the nuance of a personal trainer's feedback, making every workout feel even more personalized and rewarding.

What it does

Upon logging in, users have dual pathways:

Manual Upload: Users can introduce their .gpx file via a dedicated upload interface.

Strava Sync: Alternatively, they can synchronize with their Strava account. Once done, Rhythm autonomously generates nuanced comments for their most recent workouts.

How we built it

For the frontend we used React and material UI to create a user-friendly and appealing interface for our product. Firebase serves as the auth provider and Firestore as the database. On the main page all the comments displayed are fetched from the users collection of comments in firestore.

The backend is written in python with the Flask library. We created an endpoint to take a .gpx file and process the data, get a comment, and add it to the database. This ties together several API's and produces our end product, workout specific feedback and comments.

Challenges we ran into

  • Prompt-engineering our GPT model to output in a style and format that satisfies how we want it to be.
  • Figuring out how to seamlessly integrate Terra API to our pipeline in order to get better results

Accomplishments that we're proud of

Beyond building a holistic product, we've crafted a versatile tool with potential integrations across an array of sports diary providers – from industry giants like Apple, Garmin, and Polar to niche platforms. The feedback generated can be a valuable addition to any fitness platform, making workouts more insightful and engaging.

What we learned

  • The intricacies of integrating multiple APIs into a unified pipeline
  • The nuances of prompt engineering to fine-tune AI outputs
  • The importance of user-centric design in product development

What's next for Rhythm

  • More data. We want to train our model on more data to improve the quality of the comments.
  • More integrations. We want to integrate with more fitness platforms to make our product more accessible.
  • More features. We want to add more features to our product to make it more useful for our users.
  • More users. We want to get more users to use our product and get feedback from them to improve our product.

Built With

Share this project:

Updates