New York City may soon undergo a major technological revamp. Earlier today, the mayor’s office announced it had begun hiring for a new group that will travel across city agencies and deploy technologists to transform often wonky and confusing digital interfaces for constituent services within a matter of months. The goal: make dealing with the city government a far more pleasant experience.
Leading the effort is Lisa Gelobter, the city’s new chief technology officer and the commissioner of the city’s Office of Information Technology. Gelobter was previously involved in creating the US Digital Service, which began during the Obama administration and helped launch Healthcare.gov. Now, she says she’s trying to bring the values of that group, and similar ones throughout the federal government, to the five boroughs.
The new team, which is called the Public Interest Technology Crew—affectionately known as the PIT Crew—will begin its work by supporting Click-to-Cancel protections, a new citywide rule that bars businesses from making it difficult for people to cancel subscriptions. The group plans to build a portal where consumers can quickly file complaints against companies that they allege are not compliant. The group, which will include engineers and project designers, will soon take on four more projects, with some support from the Rockefeller Foundation.
“The idea is basically a small group of technologists working on a shorter life cycle project, right? These are not the kind of massive one-year, two-year, 10-year projects,” Gelobter told Fast Company. “These are 12-week projects, soup to nuts. Discovery to roll out.”
This interview has been edited for clarity and length.
If you were going to diagnose the state of New York City’s IT, how would you describe it?
In terms of the terminology that I use, I use the word technology, rather than IT. I think IT is one portion of the things that we do, and it’s a very critical portion. But the PIT Crew,I wouldn’t call IT. One of the things that I want to try to to shift—conceptually—is that tech is not just operational execution. It is actually a strategic asset. This is just one example of technology being a strategic asset for use in government to serve the administration’s priorities, as well as really use that to serve New Yorkers the best that we can.
When considering technology as such a strategic asset, are there agencies that are doing well at this? Are there programs that you think are really like exceptional? Others that you kind of really like think deserve more love or more care?
It depends. The challenge with technology in government is that different teams have different mandates. Because of that, they also have different restrictions. I fundamentally believe that all the technology and organizations across city government are all, in fact, doing the best that they possibly can. They are all smart. They are all dedicated. They are all in service of their of the mission, either of the city or their agencies. So, like, there’s no ding for me on any of those things.
Now, a dev team that is set up to do a citywide procurement software is very different from a team that is the PIT crew that’s going to do things in 12 weeks.. They all have a place, and they are all needed.
I know that an experimental chatbot that got started under the previous administration was shut was shut down. Do you see a role for AI, here? When I say AI, I really mean like large language model technology. Is that something you’re looking at as part of this initiative?
PIT Crew is not specifically an AI project or an AI initiative. I think with everything that we do, we are always looking at the best tools for us to utilize in the delivery of our project. Is user research the right methodology? Are we making sure that we’re going out and we’re doing design thinking?.. AI could be considered as a tool as we build out the services, but AI is not one of the primary drivers of what the crew is doing or how it’s doing it.
.How do you see the balance between just building something from scratch, doing something open source, just buying software in some cases, and just sort of like biting the bullet and procuring the tech you need?
We don’t come in presupposing the answer. That’s the wrong way of doing things. … Once you’ve defined what the problem statement is, what success criteria looks like, then you take it from there [and ask]: how do we achieve that? We’ll do the first one, with Click-to -Cancel. What do we learn from that? Okay. What how do we what do we need to change? What do we need to alter? Can we build on that?
