PocketTrainer.ai Pitch Deck: https://docs.google.com/presentation/d/1PCfEOloPKSGqBLzznd1GimZqWRPkw09puxeSZu3gb-s/edit?usp=sharing

Inspiration - Running into the issue of not knowing how to improve our workout form at the gym

What it does - Analyze your form and provide feedback on how to improve your form (in the future). The current solution just classifies workouts.

How we built it - Quick deploy using no-code through bubble.io and peltorion to get the product to the market and get feedback.

Challenges we ran into - Definitely training our model, integrating it with bubble.io. Also being able to switch pages on a 'native' webpage.

Accomplishments that we're proud of - having a team of 3 non-developers, that learned about AI, front end, and pitching in a week.

What we learned - A lot of the terminology and high-level mathematics behind AI. Converting MP4 (videos) into usable data for an AI model to be trained on.

What's next for PocketTrainer.ai - Analyzing users' workout forms in real-time and telling them how to adjust their limbs (ex. hips, elbows) to improve their workout form. Also creating a social eco system with social accountability through in-app groups and betting to see who reaches their weekly fitness goals. Lastly, have a workout feed (similar to Instagram) with your followers.

Built With

  • bubble.io
  • gcp
  • peltorion
Share this project:

Updates