I am just an ordinary guy with a pretty simple daily routine. That is, work, hit the gym, come home, watch anime and sleep. That's pretty much my life on most days, and honestly I enjoy it. Ever since my college days, going to the gym and watching anime have been my two favourite ways to unwind. Whenever work gets stressful or I have some free time on my hands, I will watch anime. Yed, I am an anime nerd, and I have no complaints about that.
Among the anime I've watched recently, Solo Leveling has been one of my favourites. I enjoyed it so much that, after finishing the anime, I actually picked up manhwa to find out what happens next. It was the first manhwa I ever read, and I got hooked up on the story pretty quickly. What I liked most is watching how Sung Jinwoo grow stronger from being the weakest, level up and take on immensely powerful opponents through the story.
What mostly caught my attention was Jinwoo's level up system. I've used multiple gym tracker apps to keep track of my workouts, but the idea of turning fitness into something like a gamified level up system really appealed to me. So, I started searching online for apps that offered a similar experience. I've found quite a few apps inspired by the same concept, but most of them felt like ordinary todo apps with a gamified layer on top or way too expensive for me to afford. After trying to find something that suits my needs, I found none of them quite offered the experience I was looking for.
So, I decided to build my own app from scratch.
I've learned Expo before, but never really got the chance to use it in a real-world project. I thought this would be a good opportunity to put that knowledge to use. My plan was to build the app that way I wanted it, using Expo and SQLite without relying on a backend server. Everything would run locally on the device, so I can use the app even without an internet connection.
I knew I could do it myself, but there were two problems. First, my experience is primarily in web development, and mobile app development is still new to me. Second, time was a concern. I hardly get time to work on side projects due to my busy work schedule and I already have some projects on priority, and I couldn't afford to spend too much time on building a fitness app.
Besides, I didn't need a perfect, production grade application. I was building primarily for my own use, and I planned to open source it so others could benefit from it too.
So, after weighing everything, I decided to take a different approach. I've decided to vibe code the entire app from scratch.
This was the first project I've decided to build entirely from scratch using AI. I started by using Claude to generate the initial version of the application. At that point, my requirements were pretty simple, I wanted to add my own workouts, build custom workout plans and have a level up system to track my progress.
The version Claude generated wasn't perfect, but it worked well enough for my needs. The problem was that I am using Claude's free tier, and the usage limits made it difficult to keep modifying the app and adding features the way I wanted.
That's when I came across OpenCode, which allowed me to use the Big Pickle model for free. I decided to give it a try, and I've been using it to and improve the application ever since.
What started as a simple fitness tracker has gradually evolved through multiple iterations. As of writing this post, the application has reached version 2.5.0.
The project is open source, so feel free to give it a try. I'd love to hear your feedback, hear your ideas for new features and welcome contributions from anyone interested in making the app better.
Check the source code on github: https://github.com/ai-craftworks/beyond