Most AI projects never make it past the demo. I build the ones that do.

Fazal took a vague brief about automating public-space operations and turned it into a system we could actually put in front of government stakeholders at GITEX. He thinks about the operations side that most developers ignore.

I hired Fazal as a senior developer and he left as the person the whole mobile team ran through. He built the QA process, the hiring process and the release process — none of which I asked him to do.

We needed a Flutter lead who could work across three time zones without hand-holding. Fazal shipped, communicated clearly, and pushed back when our specs didn't make sense. That last part is rarer than it should be.

He told us AI was the wrong solution for two of the three things we asked about, and built the third one properly. I've never had an agency turn down scope before.

I run a small distribution business and I don't code. His Urdu AI course was the first thing that explained automation in a way I could actually use the same week. We now handle orders on WhatsApp automatically.

Fazal reviewed our architecture before we raised, found two decisions that would have cost us a rewrite in year two, and gave us a build-vs-buy call we could defend to investors. Three sessions, enormous value.

I joined his team as a junior with six months' experience. He reviewed every pull request I opened for a year and explained why, not just what. I lead my own team now because of that.

Working with government departments is slow and Fazal never treated that as an excuse. The traffic police app went live on schedule and is still running years later.