QanoonX didn’t come from a market analysis. It came from watching a junior advocate lose a Saturday to bookmarked PDFs.
One of our co-founders has a brother at the bar in Islamabad. He’d watch him prepare petitions the way most people prepare taxes — with seventeen browser tabs, three photocopied volumes of PLD, and a spreadsheet of citations that never stayed current. For a profession whose entire output depends on precedent, the tools were astonishingly bad.
What we actually built
QanoonX is, at its core, three things:
- A retrieval index of 50,000+ Pakistani case laws, spanning superior courts, tribunals, and statutes.
- An LLM layer that answers legal questions by quoting — not paraphrasing — the cases it relied on.
- A bilingual interface that respects the fact that real legal work in Pakistan happens in both English and Urdu, often in the same sentence.
Why this first, and not something with a bigger market
It’s a fair question. The Pakistani legal software market is tiny compared to, say, a global CRM. But the problem was real, our proximity to it was unusual, and the quality bar for grounded AI in regulated industries is a bar we wanted to earn. Build the hard one first, and the easier ones feel easier.
Two years in, QanoonX has 10,000+ users. More importantly, it gave us the muscle for everything else we’ve shipped — PawnCRM, TeamContext, and the custom product work we now take on for clients.
If you’d told us in 2022 that our first hit would be “legal search in Urdu”, we wouldn’t have believed you. Build what’s in front of you, not what’s in the deck.