I built Training Tracker Classic - v1.0 - during my freshman year of college. I was 18, working as a Team Leader at Chick-fil-A (where I still work today), and I had just enough responsibility to see a real problem: there was no good way to track how new hires were progressing through positions like iPads, Register, Drive Through, Staging, and more. I figured I could build something. I had barely touched HTML and CSS at that point, but I was determined to learn as I went.
The database situation was bad. I was too intimidated by SQL to even try it, so I stored everything in a JSON file on the server. Every team member read from and wrote to the same file at the same time, with nothing handling conflicts. I did not know what a race condition was when I wrote it. I found out the hard way.
I hosted it on a free tier of Glitch, which shut the server down after a few minutes of inactivity. Every time someone opened the app at the start of a shift, they had to wait through a cold start. There were no user profiles, no login system, and no access control. Anyone who had the link could go in and change whatever they wanted.
I never built a mobile layout. I designed it on a desktop and assumed that was good enough. It was not. The whole thing used floats, hardcoded pixel widths, and margins I had copy-pasted from a tutorial without fully understanding what they did. On a phone it was basically unusable.
It took me months to put together. This was before AI tools, before GitHub Copilot, before anything that could answer a question in seconds. When I got stuck, I dug through Stack Overflow and rewatched YouTube tutorials until I found something that worked. I had no one to ask. Every small thing that worked felt like a big win.
We tried it in the store and it fell apart almost immediately. It could not handle multiple people using it at the same time. It was too slow to be useful during a busy shift. Nobody wanted to find a laptop just to update a training record. I pulled the plug on it not long after.