building my first expense tracker

During the period when I was interning at HeyMax, I was also exploring the idea of building a fully functional app of my own.

At that time, I was already tracking my expenses and had tried a few different expense-tracking apps. None of them really felt right. Most mobile apps only worked well on mobile, and whenever I was on my PC, I couldn’t easily key in expenses unless I picked up my phone.

When I was in university, I was on my laptop most of the time. I really disliked having to switch to my phone just to log expenses. Of course, you could argue that I could just key them in immediately after paying, or later when I was on my phone anyway and that’s fair. But putting that aside, it still didn’t fit how I worked.

Because of that, I decided to start using Google Sheets to track my expenses instead. It worked well on desktop, but on mobile it was a terrible experience. Google Sheets is clearly not built for mobile input.

That was when I started exploring the idea of using a Telegram bot as a middle layer to interact with Google Sheets. I had some prior experience working with Telegram bots back in polytechnic, so this idea made sense to me. I was also on Telegram all the time, and Telegram worked well on both PC and mobile. At that point, it felt like a perfect fit.

This was when I started working on the bot seriously looking into Google APIs, figuring out how to store data in a database, and how to sync everything properly with Google Sheets. This was the first time I worked on a “big” project outside of a school setting.

I started with Google Firebase since I needed access to the Google Sheets API. And honestly, back then AI wasn’t really a thing yet. It was just me trying to figure everything out on my own from understanding how Google Sheets worked, how the APIs behaved, and how I could make the whole experience simple and usable. I spent around three months trying to get it right.

Eventually, I decided to share it and launch it.

The response was good, but there were also obvious issues. Since I was the one who built it, I knew exactly how to use it. For new users, it could be confusing. I tried to fix this by adding tutorials, filming YouTube videos, and writing guides to help people get started.

Around the same time, I was also filming YouTube videos and sharing finance-related content for students. I noticed this was a market that felt underserved, and it was also the period where I really put myself out on social media and got a first taste of what it felt like to ship a product publicly.

I did think about monetising TeleFinance, but I never really figured out how to do it properly. When I built it, my main goal was simply to make something that worked for me. Expanding it to other people was more of an afterthought.

In the end, TeleFinance was something I personally used for about one to two years. Over that time, I gathered feedback from users and learned from my own mistakes before eventually moving on to work on something new.