-
-
Choose your target muscle
-
Choose your time
-
Choose your area
-
Choose your level
-
And get excercise
-
-
Inspiration
I often found that choosing the right workout takes longer than the workout itself. Apps felt generic, and finding proper form videos was frustrating. I wanted to build something smart and simple that works like a personal trainer but instantly and for free.
What it does
GymGenie is an AI-powered fitness assistant that generates a custom workout plan in seconds. I just select my target muscle, time, fitness level, and location, and it creates a structured routine with warm-up, main sets, and cool-down. Every exercise comes with an embedded YouTube video so I know exactly how to do it.
How I built it
I used HTML, CSS, and JavaScript for the step-by-step user interface.
A Node.js + Express backend calls the Groq API to generate workout plans.
The YouTube Data API pulls and embeds the right exercise videos automatically.
I focused on making the design minimal and beginner-friendly.
Challenges I ran into
Getting the AI to output a consistent format that my frontend could parse.
Handling exercises that didn’t have pre-mapped videos.
Balancing features while keeping the flow simple and not overwhelming.
Accomplishments that I'm proud of
Building a fully working AI fitness assistant in hackathon time.
Successfully integrating real exercise videos into the generated workout plan.
Creating a clean and intuitive flow that feels like a personal trainer session.
What I learned
How to craft better prompts for structured AI outputs.
How to integrate multiple APIs into one smooth workflow.
That simple, useful features matter more than cramming in too much.
What's next for GymGenie
Add progress tracking so users can log their workouts.
Build a mobile app version with notifications for consistency.
Enable voice commands for hands-free interaction.
Connect to wearables for real-time, data-driven workout suggestions.
Log in or sign up for Devpost to join the conversation.