I decided to review my old projects... I was shoked!

Othmane CHAOUI — Sun, 01 Nov 2026 21:53:55 +0000

For the past three weeks, I've made a bet with myself.

Not out of motivation, but rather because of the pressure coming from the current state of the software engineering job market and the uncertainty of finding an internship by 2027.

The bet was simple:

Every single day, I had to make at least one meaningful push to GitHub.

Not pushes containing school practical work or exercises just to make my GitHub contribution graph greener (I know GitHub changed the colors 😅).

I would primarily focus on my old projects: reviewing, fixing, and improving them.

And honestly, I was shocked.

Shocked by how blindly I used to trust AI, copying whatever it generated straight into my projects without properly verifying or validating it.

I started discovering serious problems in my projects' overall structure and architecture:

Business logic handled by completely irrelevant files.
Functions placed where they didn't belong.
UI files directly accessing the database.
Poor commenting practices.
And many other architectural and design issues.

Some projects even had to be almost entirely reconstructed because of these problems.

What struck me the most was realizing how immature my previous coding practices and project structures were compared to how I approach software development today.

I never thought that just two years in the field could make such a difference in my knowledge and engineering mindset.

I guess the hundreds of software engineering articles, discussions, and videos about best practices, conventions, and architecture really did their job.

More importantly, I've become much more conscious of how I use AI: not as something whose output I blindly trust, but as a tool whose suggestions I need to understand, question, and validate.

Sometimes, looking back at your old code is the best way to measure how much you've grown.

DEV Community: Othmane CHAOUI

I decided to review my old projects... I was shoked!