The first thing I will work on


The shittiest ideas you can find

This week I set myself the goal to find at least 10 business ideas to work on. I found 14 problems that I thought might be worth solving. After looking onto them after a few days, most of them are trash. But there are two sticking out.

Let's explore them.

1. Applications have bugs, every app, every software

One of the biggest issues we face in software development are either related to bad project management, or issues introduced by a colleague or oneself. Bugs are tedious. Discovering bugs is work. Fixing them is even more work. And making sure they won't surface again is another thing.

That's where I see a few opportunities. I see this in my projects everyday, we do have a few test automation engineers, mostly from the Ukraine. They do speak fairly well English, but surely not at a level that enables them to write decent bug reports, or even understand complex business cases. The app we're currently building for the client is a SaaS product, yet without any error reporting and capturing.

My idea is to have something like that captures frontend facing errors, but also backend errors. The product collects information about the stack trace, or rather how the state was produced. These information are used to generate bug reports into a tracking tool like Linear or YouTrack. With this bug report a developer can tackle the bug easier.

At a later product stage, an AI could generate regression tests based on these cases. Either Unit or E2E tests.

This product definitely fulfills the commandments of the CENTS frameowrk.

2. A chrome extension that fixes an annoying issue with ChatGPT

Have you ever asked ChatGPT to write a function for you, and it somehow started to output some parts, but very often didn't gave you the full function? Telling you to implement the function? Like it's lazy. Having a simple Google Chrome Extension to automatically parse the output, discover these parts and auto prompt it to complete these functions / write them out for you.

This is something that does not really fullfill the commandments of the CENTS framework to be fair. I wouldn't know how to price this.