If this is what it takes for Dallas, what does it take for everyone else?

I spent Thursday night at a Dallas AI meetup listening to the person who leads data and AI work for the City of Dallas walk through roughly a decade of city projects. It wasn't a deep dive into one system — it was a tour, and that's what made it worth writing about.

A decade of small, unglamorous fixes

Over the course of an hour, the talk covered a model that scores which buildings are most likely to catch fire, so inspectors know which ones to check first. A text-mining tool that reads paramedic and police narratives overnight — a job that would have taken one person 7.6 years by hand — to flag children at risk of trafficking. Cameras on sanitation trucks that spot illegal parking and closed-park violations. A chatbot for zoning questions across 1,500 planning districts. Drones dispatched to structure fires and active-shooter calls, with a human always watching the feed.

None of it looked like the Hollywood version of a city "using AI." Most of it was: read the reports faster, connect two databases that don't talk to each other, tell a person where to look first. The value was almost always in reducing wait, not replacing judgment.

The speaker described spending time in the field — riding along in squad cars and fire trucks, joining building inspections — to see what the frontline work actually looks like before building anything for it.

The question I couldn't resist asking

During Q&A, I asked about the city's permitting intake process — whether the backlog was in the review itself or in the back-and-forth before it. Turns out Dallas had already worked through it: median time for a single-family permit dropped from 80 days to 8, and the commercial backlog is down roughly two-thirds. Most of what's left, the speaker said, is exactly the back-and-forth I'd guessed at — incomplete applications bouncing between an applicant and a reviewer.

It's a small, unremarkable-sounding fix. It's also the kind of thing that changes how long someone waits to get a roof repaired or a business open, which is exactly the kind of impact this column is here to notice.

What that means for everywhere with less

Here's the thought I left with. Dallas is the 9th-largest city in the country. Its public safety budget alone is over a billion dollars. And by the speaker's own account, the team is small, resource-constrained, and still working through a backlog of "greenfield" — basic, unglamorous data work that hasn't been done yet. One of the team's signature tools started as a side project, coded in a school cafeteria.

If a city with that much budget and that much talent is still doing this work one small team, one use case, one detective's frustration at a time, it raises an honest question about everywhere with less: smaller cities, under-resourced counties, entire countries without someone in that seat or a budget line for one. The tools that recovered 77 kids in a year and cut permit times by 90% aren't exotic. They're achievable. The gap isn't the technology — it's whether a place has the people, the time, and the mandate to build it.

That's the part of this work I keep coming back to. The human and social impact isn't only in what gets built. It's in who gets left waiting because nobody had the capacity to build it for them yet.