Building a
career that
wasn't supposed
to exist

When I started looking into AI governance and legal technology, I kept searching for people who had already done it in Pakistan. I wanted the checklist. The guide. The sequence.

I found almost nothing.

No senior from my university doing exactly this. No established pipeline. No one saying: start here, do this, avoid that.

That absence turned out to matter. It meant there was no orthodoxy to defer to. No conventional sequence to follow. The work I've done — at Lex Lata, on The Digital Gavel, on the Land Revenue Act chatbot, on this site — all of it started as a kind of improvisation. Not recklessness. Improvisation. The difference is that improvisation still has intent.

Law taught me how institutions think. Data science taught me that institutions are often making decisions about systems they barely understand.

The more I studied both, the harder it became to stay in only one. Law without technology felt incomplete. Technology without governance felt dangerous. Policy discussions lacked technical grounding. Technical discussions ignored legal consequences.

I became interested in the space between those conversations. That space — where regulatory text meets model architecture, where constitutional rights meet algorithmic outputs — is where most of my work now lives.

It's also still underbuilt. Which means there's room.

My career strategy looked less like a roadmap and more like opening twenty browser tabs and refusing to close any of them.

If someone looks at my CV quickly, they might see internships, research projects, certifications, publications, scholarships, competitions — and assume there was a grand strategy behind all of it.

The truth is less glamorous. A lot of it came from curiosity. A lot came from saying yes before I felt ready. Some applications were rejected. Many emails were ignored. Some projects quietly failed.

Most people only see the successful version of a journey because failure rarely gets a LinkedIn post. This page is trying to be a different kind of document.

One of the strangest parts of this journey is that I've spent most of it building things before feeling qualified to build them.

The website. The writing. The Digital Gavel. The legal-tech experiments. The research at Lex Lata. None of them arrived after confidence. They arrived before it. The confidence showed up later — sometimes much later — as a consequence of having shipped something real into the world and had it tested by actual stakes.

Working on live IHC filings calibrated my attention in ways that academic exercises never could. When a wrong citation has consequences for a real client, precision stops being a virtue and becomes a minimum.

The cross-jurisdictional work — Pakistan, India, UAE, US legal frameworks — started as a byproduct of different opportunities. It became deliberate once I realized how rare it was. Most people in AI governance come from a single legal tradition. Someone who can map GDPR onto Pakistani data protection bills, or connect an Indian court's reasoning on algorithmic accountability to EU precedent — that's a much smaller group.

The South Asian regulatory context is underrepresented in the conversations that matter. I treat that as a positioning advantage, not an obstacle.

Four things. Specific ones, not general wisdom.

What keeps me going isn't ambition. Ambition sounds neat. What I have is closer to intellectual restlessness — a set of questions I find impossible to ignore.

I don't know. But I intend to find out.

How to scale the work without losing quality. How to position for opportunities that don't exist yet in my immediate geography. How to build institutional presence that makes research matter beyond the people who already follow it. These are still open.

This page will stay honest about what I know and what I don't.