I went to school with the son of the owner. For about 3 years off and on he'd been asking me about building apps, where he could find good devs, what price to pay.
I'd sent him some info, but never got too involved. Then the pandemic hit, and I found myself with a bit more time to kill, so I listened more to what they needed.
They basically wanted to get a loyalty program going at their gas stations and partner gas stations, but wanted custom rules which their current provider couldn't do with their off the shelf solution.
They also wanted to use one of those old school CC machines that were everywhere before the likes of square showed up. But getting your hands on those units is expensieve.
So I started planning it out. They needed to store customer data, and needed a point of sale device to swipe customer loyalty cards on, which needed to read in customer info, do a deduction, or a store credit top up and print a receipt.
I found a android terminal machine off alibaba, opened up android studio and started hacking away. I hadn't done more than a basic android hello world at this point.
For the backend I used django to get started quickly, and they were very happy to use the django built in admin panel to do stuff as needed, and view transactions.
After I got the devices talking to django, I had to integrate a loyalty vendor they already sourced, and just went along with it, but that was a mistake. The vendor points API doesn't really add much value on what django can already do for me, and my new goal is to become that vendor since I think I can do their product better.
But basically I iterated on it for about a year, launched a little over a year ago, still working out some edge cases and kinks, but they do roughly 25K transactions a month on devices spread out at over 20 stores.
They want to bring it to new islands, so I'm trying to remove some of my duct tape fixes with more stable fixes so that can be smooth sailing come next year.
EDIT: Might get back to writing and update with a much longer post. Draft: