![Podcast microphone against a blank background](/sites/g/files/omnuum10826/files/styles/hwp_21_9__1920x825/public/datasmart/files/podcast_microphone.jpg?itok=3Q9qYvou) # How Cities Scale Data Innovation Through Cultural Change Episode Ninety-Eight July 29, 2026 [ Betsy Gardner ](/people/betsy-gardner) Data innovation isn't about buying better tools. As Dallas and Cleveland show, it's about how leaders listen, support risk-taking, and work alongside agencies. The result? A culture where innovation thrives, and scales. Host [Stephen Goldsmith](/stephen-goldsmith "Stephen Goldsmith") speaks with [Dr. Brita Andercheck](https://dallascityhall.com/departments/Data-Analytics-Business-Intelligence/Pages/director.aspx), Chief Data Officer in Dallas, and [Dr. Liz Crowe](https://www.clevelandohio.gov/city-hall/departments/innovation-technology), Chief Innovation and Technology Officer in Cleveland, about how they drive transformative change across city governments. From reporting structures to how you listen to agencies to going out into the field yourself, they share the unglamorous work that actually drives transformation. **In this episode, you'll learn:** - How to strike the balance between centralization and decentralization - Why a problem-first approach beats technology-first - Why data teams should measure their ROI - The importance of collaboration and relationships - How to create psychological safety so teams "fail smart" **Guest:** - **Dr. Brita Andercheck** – Chief Data Officer and Director of the Office of Data Analytics and Business Intelligence, city of Dallas - **Dr. Liz Crowe** – Chief Innovation and Technology Officer, city of Cleveland Embed *Listen here, or wherever you get your podcasts. The following is a transcript of the conversation.* **Steve Goldsmith:** Thank you, and welcome back. This is Stephen Goldsmith from the Bloomberg Center for Cities with another one of our podcasts. Today, we have two of the most influential city leaders in the country, particularly on the relationship between data and innovation and change. This is going to be just a terrific conversation with Dr. Brita Andercheck, the chief data officer and director of the Office of Data Analytics and Business Intelligence in Dallas, the longest title in America, and Dr. Liz Crowe, we have a lot of doctors on the call, chief innovation and technology officer in Cleveland. Welcome to both of you. **Brita Andercheck:** Thank you, Steve. Excited to be here. **Liz Crowe:** Thanks, Steve. Excited too. **Steve Goldsmith:** Good to have you. Well, I've worked with both of you on and off and been influenced by your perspective as well as your intelligence and drive, but let's just start a little bit with how you both got to your jobs. Brita, you're higher on my screen, so I'll just arbitrarily pick on you first. **Brita Andercheck:** Happy to jump in. So I'm an academic by training, as you mentioned. And so, I spent a lot of time in school preparing to be a professor. I started my career on faculty at Southern Methodist University, and I loved teaching. I loved it. And I was a couple years in, and we were teaching... It was a course on social problems, and my students, quote, unquote, "solved the social problem of homelessness in the city of Dallas in a neat 45-minute seminar." Right? And so, as I'm leaving class and I'm walking across the beautiful, leafy green quad of SMU, I'm thinking to myself, "Gosh, shouldn't I actually try to solve that problem as opposed to solve it in the abstract in the ivory tower? Shouldn't I go get my hands dirty?" And I knew that I didn't know anything about solving homelessness. I had no subject matter expertise in that area or in city government, but I did know that I knew how to measure things and iterate. Right? And so, that brought me to the city of Dallas where I joined the transportation department. They had a lot of data issues that they were focused on around crashes, solving crash problems. A bad accident means somebody doesn't go home to their families, and it's a huge impact. So we started there. We started by getting over $100 million in grants from the federal government to rebuild signal intersections, and that kind of set us on a trajectory. The city got really serious about data and the results that we could drive with really good data when applied in specific places. And so, I was in an operational department, and that meant that I knew the demands when you're providing services to residents. That is how I ended up here, and the city manager said to me one day, called me up to his office, and said, "Brita, I want you to do what you're doing in transportation citywide, and I'm going to give you a department. I'm going to appoint you the director. I'm going to provide funding." And I remember just sitting there thinking, "How is this even going to work?" And I went on for an hour about all the reasons it wasn't going to work, mostly about I didn't know how I was going to get IT to give me the equipment I needed, and I didn't know how I was going to get this or that done. But he persevered through my resistance and concern about it not working and said, "Let's do this thing." And it's been six years. So it's been exciting. **Steve Goldsmith:** Well, that's a great background except for the implicit insult to those of us who are merely academics and don't actually accomplish anything. I'm so glad you mentioned that... **Brita Andercheck:** No \[laughing\] **Steve Goldsmith:** ... you were just merely working in a university and really not getting anything done \[laughing\]. So I'm going to hope that our next guest, Liz, can you be a little more complimentary in your background? **Liz Crowe:** I consider myself a recovering academic, because I started my career in academia, went and got a PhD in public administration and policy, have always loved the public sector government space. I lived and worked in D.C. for a while, kind of did a bit of the federal government thing, and found myself back in Cleveland, and was here for a number of years when there was a mayoral transition in 2022. So Mayor Bibb took office, and there was really a full realignment of leaders across city government. And I remember it distinctly, he posted looking for someone who could lead an analytics team, because they realized that we did not have a data practice. In fact, the talk track within City Hall was, "We don't do data." Multiple staff told me that, and Mayor Bibb came in, and he's like, "Well, where's my dashboard?" They're like, "Here's your bar chart." He's like, "That is not what I was talking about." And I saw the posting. I remember I sent it to my sister, and she was like, "That's you. That's absolutely you." And I arrived here to truly a greenfield development in data. So we had no analytic infrastructure. We didn't have any policies. We didn't have a program. We figured out we actually had a Microsoft license with some Power BI licenses already in it. So that was cool. That was a good first step. But we really had to start at ground zero and start not only talking to our staff about data, but also start talking to our technical people and started hiring in a technical team who had a data practice and data best practices. At the same time, I was also handed the transformation of our 311 team, which really gave me an in-depth look on the operations of the city, which was pretty cool, and getting to understand, "What does public works do? What does parks and rec do? What does building and housing do? How do they operate? How do they track their work in our systems?" And that started to build out reporting and analytics and all the components that go with it. **Steve Goldsmith:** The job of change agent can be really located anywhere. I want to have you both reflect on why you're so successful. And Liz, I've written about this a little bit with you, and I experienced it when I was deputy mayor of New York. If you centralize expertise too much, you lose the innovative juices from the agencies. If you totally decentralize it, you lose the ability to drive new technologies into the agency. So talk a little bit, maybe you first, Liz, and then Brita, about, who do you report to? How do you instill the appetite for change? And how do you interact with the agencies to produce transformation? **Liz Crowe:** Yeah. So I report directly to our chief of staff, and we have since absorbed our analytics function as an office under the Department of Innovation and Technology. So we've got IT, 311, and then Urban Analytics and Innovation as three offices, meant to be sisters, not to compete with each other, but they all have really different purposes. But our data work really started out as a mayor's office. The first thing that came to mind is I like to tell my staff, "My job is to be delightfully obnoxious." So I am a total champion for this work. I will talk to anybody. I will buy anybody a cup of coffee, understand where they're coming from, figure out where they are, and then really try to set the North Star for where we're going. And oftentimes, I find many of our city workers, they've been to the conferences. They get the concept. They've seen the cool presentation. They've seen the vendor demos. They know. But then they get back to their desk, and they have absolutely no idea how to make that thing happen. So what they need is they need sort of a central hub with expertise who can convene and guide and lead and knows how to haggle with the vendors and all that kind of stuff, who is also friendly and approachable and knows how to talk to people, and then we have to house that expertise with the business, because this is where you need the owners and the champions on the ground who understand the tactical work, and you need a leader who's able to say, "Hey, this is what we're going to drive toward." And just start clearing brush to be able to get the team moving along that path. **Steve Goldsmith:** So, Brita, I remember you telling me an earlier point. You buy people lots of cups of coffee as well. So I now appreciate that, but how do you cause so much change? I mean, you both are very unusual in your ability to drive change. So how do you power change in the organization? **Brita Andercheck:** Yeah. First of all, that's a great question. And obviously, Liz and I power it with coffee. Right? That's clearly one of our secrets. So let me just say, I report to the CFO, who reports to the city manager. Finance is sort of an interesting place to place data analytics, but one of the drivers for that was we wanted to ensure that, A, the office was funded, but that also that it was uniquely aligned with budget and performance. We really wanted the efficiencies being driven out of this office. So that was sort of the initial alignment that we have, and we're not connected to ITS. So how do we get people to do things right but without driving too much change and without just telling them what to do? And I think that the story that I might have shared with you, Steve, was that no agency head has ever called me and said, "Brita, I need a linear regression." Right? But they'll call me and sit around and talk about the problems that they're experiencing in their operations. Right? I was hosting an event for grad students at City Hall last week, and one of the students asked me a very similar question. They said, "How do you know what the problem is you actually need to solve?" And I said, "I also ride in squad cars, and I show up to fires, and I go on building inspection visits with building inspectors, and I watch what they're doing." Right? And so, I think it's really important that you're paying attention and listening to whoever that frontline person is. Who's the end user of the technology or the data product you're building? Because they truly are your most important customer. You can build something fabulous that's perfect that nobody ever uses. And so, I think it's really isolating who is the person that is actually the client. We call them clients, the people in the agencies, the patrol officers, whoever it is that's going to be using the thing you're building. Are you actually solving a problem that they have every day? And I think you have to get comfortable with the fact that people have been to the conferences, as Liz said, and they know that they need something, and sometimes they're attached to an idea, a specific idea. People give me a lot of ideas, but what they really want is a partner in helping them make positive change. And so, there's a lot of stuff that I let go in the beginning, because I know by the end, by the time we get to the end of this, they're going to be thrilled with the product, and they're just going to be happy that somebody came alongside them and supported them in their mission. **Steve Goldsmith:** I'm particularly interested where the technological innovation also results in business process reengineering. So it has an agency effect and has a technology effect. And so, there's all these old ways we issue tickets for graffiti and code enforcement and repair streets. So you both are experimenting with cameras and how that might change work processes. Talk to us a little bit about your camera experiences now as an example of change that requires you and the agencies to work together on process change and technology change. **Brita Andercheck:** Yeah. So we are working right now. We've got cameras on our sanitation brush trucks. We chose those trucks because they had a good kind of overview of the city, and we're using those cameras to detect high weed, code violations, properties that are in blighted condition, graffiti, et cetera. Things that used to take a code officer going out in the field and finding or a neighbor proactively reporting, we're trying to proactively capture on video feeds. So there was big concern, "Well, if we adopt this and we increase the number of citations we're writing, how will we ever go out and have enough code officers look at all these?" So one of the things that we do know is that when we issue letters as the city, we get about an 80% voluntary compliance rate. So we know that if we write a letter to a property, there's a pretty good chance that they're going to make the change. So even though we know that we're going to be generating tens of thousands of more potential notices of violation, our hope is that our voluntary compliance rate remains high. But also, we'll be able to have that video feed to go back and say, "Okay. We wrote you this letter. You didn't respond, and now we've got a year worth of evidence that now we can hand that over to community prosecution if we're not getting compliance." **Steve Goldsmith:** Brita, who came up with that kind of change from traditional code enforcement as, "I want to see how many tickets we can write," as contrasted to how we can create better property conditions, not necessarily more tickets. **Brita Andercheck**: The director of code brought me this idea. We had a meeting, he and myself and the city manager, to figure it out. But that was driven by the agency, and we supported the kind of change in terms of information flow. **Steve Goldsmith:** And Liz, in your situation, who owns success, and who would own failure on this particular issue? **Liz Crowe:** We are using similar camera technology for a similar issue, which is really a code enforcement-based issue, and that is going to be our Department of Building and Housing. Our leap into using cameras to start to look at property conditions and code enforcement came from actually sitting around with our director of building and housing after a 90-minute meeting where we had the IT teams and the public works teams and the building and housing teams mapping how we were going to track and issue citations for high grass on occupied properties where the city would have to cite and then, at some point, send out the public works team to cut, and then send out a notice where we would bill somebody for cutting the grass, and we said, "There has to be a better way." Over 20,000 vacant lots in the city of Cleveland, and we were staring down the tunnel at just scaled issues and saying, "Hey, this isn't scaling. This isn't working in a sustainable way. We don't have enough people, nor will we get the budget to have enough people, nor do those people exist to be hired. We have vacancies that we're struggling to fill. So what are ways where we can start to solve this?" **Steve Goldsmith:** Tell me about that design charrette. Right? You said we. Who's at the table? Who's saying, "The technology can do this"? Who's saying, "But we can't do that, because..." Just a little more about the innovation process at the intersection of technology and business. **Liz Crowe:** Yeah. I think the core of it started with, what's the problem that we're facing? And we were facing problems of staffing, problems of just the work being too much, the problems of like, "Hey, is there a better way?" And part of where we're using cameras is we are not writing citations off of what we see on a camera, but it is dramatically informing where we send our inspectors. We have a car that's mounted with the cameras. We call it our 311 City Support Car. So it's driving around. It's fully branded. So our residents know it's out there, but we'll send the car into an area. Our building and housing inspectors are now looking at those pictures first and going, "We see in advance what sort of support we would need to send in." Once we got clear on the problem, then we started to look for the technology. So we didn't go tech first. We went problem first, and that's, I think, helped guide how we prioritize the work, who's on first to deal with what issues, and then we've been able to expand a little bit. Now, our 311 team has a bounded set of to-dos that they go in and look for, things like tires, illegal dumping, where they saw it on the picture from last week. We'll put it in for our public works crew to go address that as well. **Steve Goldsmith:** And I've been in local government, for good or bad, for 30 years. So it feels to me like the problems are more obvious than the solutions. So what's your process of generating innovation? New questions, new ways to approach the issue. Right? Inherent in the last five minutes of our conversation is both of you have a new technology, but your agencies have a new way of kind of thinking about the problem. So maybe we'll start with you, Liz, and then go to Brita. How do you generate curiosity and imagination so you can come up with changes? **Liz Crowe:** I would argue I don't think the problems are always that clear, because if I put four or five people in the same room who are all from the same department or different departments working on the same problem, all of them are looking into the box from a different lens. It's like that old parable about the elephant where one thinks it's a wall and another one thinks it's a trunk, and they're trying to understand what they're looking at. And so, I think step zero is to get really clear on what the problem is and to create forums and areas where you can bring people together to map a process or have a conversation or to do that sort of thinking and give them a bit of a safe space. And then as the ideas start coming, we stack-rank them in terms of feasibility, and to have the technical people ready, not to always present the solution, but to have your technologist say, "All right. I heard you. Have you thought about it this way?" And Brita, you said something earlier of, a lot of what people come up with at the beginning may not be where we land at the end, but we're listening and learning throughout, and teaching our data and our technical people, "Hey, don't charge them with a solution. Sit back and listen. Let them haggle a little bit." And then we can start coming in with, "Ooh, did you think about it this way?" And ideas start to blossom from there. **Steve Goldsmith:** And Brita, how do you generate imagination? **Brita Andercheck:** One of the things that we've had a lot of success with is demoing things that have worked and then applying those same technologies to other problems. So a quick example. We built a natural language processing algorithm to help our Crimes Against Children Unit, and it revolutionized the timing in which they were able to get vital information about a missing child. And in 2024, they recovered 117 missing children. We implemented this tool. And in 2025, they recovered 194 missing children. Right? An increase in 77 recoveries. And once this became known to the other departments and the police department, they went crazy, and they started saying, "Well, could we use that same sort of thing for gang? Could we use that same sort of thing for robbery?" And so, one of the things that we try to do is have these sort of showcases where we talk about projects, and we let the SME, we let those subject matter experts lead a lot of that conversation, and that generates a lot of, "Huh, we know how to do this thing. We have this tool. We have this tech. What else can we apply it to?" And that's a big part of our AI strategy going forward in the city of Dallas, is we know we can do these three things in these three areas. Can we get 15 permutations of them in different departments to continue to drive those efficiencies? So I think one of the ways that you inspire imagination is you show people something that worked, and you let them know it is attainable. It's close. They just have to grab for it. **Steve Goldsmith:** I love the answer. It gives me a chance to brag for just a second. About a decade ago, when I worked for Mike Bloomberg, we were trying to set up the country's first data analytics center, probably 12 years ago, and the agencies, nobody really was a believer. So each month, we'd have the chief operating officers of the largest departments together, and two would be assigned to report on a victory using data. The goal was to kind of evangelize internally, which is kind of what you just said with your police story. Let me switch gears a little bit, Brita, starting with you. One of the challenges, I think, today is you report to the CFO. Most people in your position don't, and the return on investment calculations are not very good. So you have an idea, a big new technology, could be AI or otherwise. It costs money. You predict that you'll be successful. Nobody really believes you. **Brita Andercheck:** Well, let me say this. If you're going to be serious about this, you need to do it. If you are running a city as a mayor or city manager, you cannot think that this kind of thing is going to work if you appoint one person in one department and give them two people to help them. That is a waste of your time. If you are going to do this, do this. Jump in the water. Don't stick your toe in, and this is the future. So if you aren't doing it, you should be. To your comment about ROI, we do and we run Monte Carlo simulations every year to show what it costs us to build what we build versus what it costs a vendor to do it. And we know that on average, we are about eight times less expensive than having a vendor do the work. And on average, us just existing saves the city over $10 million every year. There are savings that can be shown statistically, that can be shown project by project. And so, we do a lot of work to show that. The other thing is, I am ruthless about what I think is successful or not. Right? So when we said we were going to try a bunch of AI initiatives, one of the things we say is we need to see a measurable benefit of this AI deployment within 12 months or we're turning it off, because we're not here to just hypebeast around AI and turn on every module and every piece of software we already own. If it's not going to be used, if it's not going to yield a measurable benefit, we're done. However, if it is going to yield a measurable benefit, where else can we apply that? How do we scale that? We have such tremendous partners in the departments that believe in us that I have more work than I could do in the next two years, Steve, in terms of the backlog of demand from the agencies. And so, this is really a balance between, which of these things are we going to do? Which are the highest ROI, and what are the political priorities in the city? **Steve Goldsmith:** Well, Liz, can you top that? That was pretty emphatic. **Liz Crowe:** I don't think I can top that, but I will say, I will echo the fact that there's a commitment that's needed. I think putting the right resources to it is really critical. And I will add, there is a certain level of superpower of bringing competency in-house and making sure we have an in-house competency. We can have conversations, city employee to city employee, that we can't have city employee to contractor, or consultant or vendor that we would bring in, because we have folks who just, they aren't sure how much of their dirty laundry that they want to air. And so, when we're able to say, "Hey, this is a competency we have in-house. We're not going away. The contract is not going to end. These are full-staffed positions," our employees are able to buy in a little better and a little faster, and we have truly been in building mode over the last four years. So we, from an analytics and a data warehousing standpoint, are just hitting that stride where we're starting to see that ROI, because we had to do a lot of stand it up. We saw a forest, and we had to go mill the lumber, and then we can build the house. Now that the house is built, we can start to really live in it in a scaled way. **Steve Goldsmith:** How do you create the environment where people are comfortable taking risk? It's easy for the three of us. Right? We have senior positions in government. We say, "Well, we have a mayor or city manager who has faith in us. We'll just take the risk." How do you create an environment where others are comfortable taking risks? **Brita Andercheck:** That's a great question, Steve. The first thing I want to say is that failure is an opportunity to learn. And so, rarely is anything a 100% failure. So there are things that we have stopped doing. There are projects that we have put on hold. There are things that we have done differently. But from that, we have learned something that became the foundation of the next iteration, or we've learned we don't want to do it that way again. One of the ways that you do that, just speaking very tactically, is they know that I'm going to provide cover. Right? I'm going to own the failure, and I'm going to give them space to fail, and part of what that means when I say space to fail is literally time in the calendar. So if this thing is due, say, to the fire chief on the first of the month, I'm going to see it on the 15th if we need to pivot. Right? And I'm going to own and manage that political pivot. So I guess failing is part of it. They say, "Fail fast." Sure. You can't really do that in government with a big public entity. We tend to test a lot of things internally before we propose that we move it into a resident-facing area, which gives us a little bit of less public failures. But we try to fail internally, and I provide that cover, but you got to have people feel they have the freedom to try things. **Steve Goldsmith:** Liz, maybe you could take your turn at that question, but let's turn it just a little bit to AI. There's a lot of big, promising things. A lot of it is bright, shiny whistles, but some of it's serious. How are you thinking about where to invest in AI as it relates to this process of innovation, change, and risk? **Liz Crowe:** Yeah. I mean, it's such a big question. And for us in Cleveland right now, we're really trying to figure out exactly where we put our AI investment, and the pendulum of technology development hasn't fully settled. So we're watching costs swing up, where they were low a couple years ago. They're getting higher. We're seeing use cases start to land, but also get riskier in different areas with folks with some stronger opinions about it. When I think about our AI strategy, we have to kind of segment things into buckets. And as we're moving through change, I will 100% agree with Brita in giving your team the air cover and putting that dome over your team in a safe space, and then, when we work with the teams on the ground, segmenting things out into pieces. I mean, we can say jump out that window and land on concrete or jump out that window and land on this soft pad, and you can say, "Hey, we're going to jump out of a lower window, and we're going to put some padding at the bottom. We're not going to go splat out of a 15-story window." And thinking about structuring projects in ways where we can make them incremental, and we can say, "Hey, we're going to get here and then we're going to evaluate, and then we're going to get the next step and we're going to evaluate." And we are just scooching ourselves along from an evaluation standpoint. When I think about AI, we have employee work assistants. So that's one category of AI, and we're working on getting those off the ground for us. We have things like cameras on the cars that are looking for high grass and weeds. That's vendor-specific data collection. That's got more guardrails, but that's also big vendor management activity, and then there's the city training our own agents. And each of those categories of AI, we're going to have to manage and structure just differently. **Steve Goldsmith:** I feel like the easy AI applications are easy, relatively inexpensive, and they're tactical. We can help every employee do their job a little bit better. I feel like many cities are too slowly approaching the dramatic changes that AI makes possible, cross-agency data, predictive interventions, preemptive solutions, personalization. You two are leaders. You're national, international leaders in the use of data for innovation. Let's make a big change in the way local government works. Increase its responsiveness. How would you do that with AI? How are you going to create that culture inside Dallas and then Cleveland? **Brita Andercheck:** The first thing you need to do to create that culture is you need to create moments of inspiration and belief, which means you need to find those early successes, and you need to capitalize on them. When something works, where else can it be used? Right? We started with a pilot in parks' cameras, which we can then turn on police and other cameras. Right? It is contagious when something works. There's an adrenaline surge when your team has a product that's successful and is reaching residents. I think the other thing to just say here, Steve, is the world of generative AI is huge. There's a lot of things that can be done. There's a lot of, frankly, snake oil salesmen still in the space, which I think that city leaders are right to be cautious of. But there is so much tremendous, relatively safe work to be done in traditional AI space of machine learning. You can mitigate risk by leaning into traditional AI, and you can turn it on right now, and you don't have to go buy a single thing. And if there's a city that's scared to start, start there. I think we all have this vision from television shows that all crime fighting happens, and there's some magical analyst. Right? Pick your favorite show. Penelope Garcia is back there with all the screens telling everyone the exact second to kick in the door. And the reality is, when you ride with a patrol officer, when you go to your crime center, even in the largest cities in this country, that is not the reality, and I'm not telling you all something you don't know. But the way you can make it the reality is you can start using traditional ML to connect this stuff, and that is going to feel like everyone's on another planet. Find those early successes. Start simple. You don't have to go out and buy everything involved. It's not like a new hobby that suddenly you're spending $5,000 to get into. You can start small. You can start in-house, and you can make a tremendous amount of change. **Steve Goldsmith:** Great answer. Liz? **Liz Crowe:** I would say it's confidence and literacy are probably the most critical from my vantage point. And that is that if our city workers understand what they're doing, they don't have to understand all the logistics of generative AI and what happens in the black box, but they have to understand, "When do I use this tool? How do I use it? What are good use cases?" And then they have to know how to explain it to somebody else, "Hey, here's how this thing works in my day." They'll adopt it. And so, as we are thinking about AI, launching it citywide, and starting to accelerate, the use cases for us are infinite, everything from call scoring and sentiment scoring in our 311 call center all the way on down to better document processing. I mean, I can come up with probably 30 of them off the top of my head just to list them out. And if the team is literate and they're confident and, like, "Yup. I know how to do this. I understand what it means, and I know the implications for my job," they will do the work, and they will deliver fantastic results. They know what they're doing. These are capable people. There is a deep, professional competency in how to run a government and how to do their work. This is just a new way of looking at it. **Steve Goldsmith:** So what does data literacy look like in a new AI world? How would you change the habit, the behaviors, the talents, the competency of your current workforce to unleash that productivity? Liz, then Brita. **Liz Crowe:** Ooh, this is a tricky one. My personal opinion is AI work assistants are going to be table stakes for office workers in the next probably three years. So right now, if you got to your office-based job, you would be very confused if they said, "Why do you need this email account and Microsoft Word is optional for our employees?" Those would be confusing things. You're going to get that suite, and these AI assistants are going to be part of that suite. And to me, from a literacy standpoint, just like an office-based worker is expected to do basic word processing, basic email etiquette, we don't even think that we're literate in these things anymore, because you just write an email, and you know how to work the mouse on your computer. You understand these baselines. And if we can get to a point where your Claude or your ChatGPT or your Copilot are just another tool in the suite that you use as part of your day to accomplish your goals, some people will be better. Some people will be okay at it. Some people, they'll be a little bumpy at it, but it is one of the expected suite. I think that's the threshold that we're trying to get to from a literacy standpoint. **Brita Andercheck:** Liz did a great job kind of covering literacy. In addition to overall employees getting used to productivity tools being a part of their new norm, I think there's also, how do you get your core data analytics team really conversant and comfortable building AI tools, building agents? That is about training and time, and really making sure that you're devoting some reasonably high-end training to that focus and then proving to them they can build it. I was at a session at the Harvard Kennedy School recently, Steve, that you were hosting, and we ended up writing some code, which became the basis of a project that we then put into place in Dallas. And that initial code, which I think I was eating a muffin when I wrote in the cafeteria up there, that initial version of code obviously got massively improved by my team over time and trained. But there is a company, a very big camera company who's got a $37 billion valuation that is asking us for it, because it works better than the one they have on board. Okay? I had a different company valued at 7 billion sitting in our Emergency Operations Center, literally saying they were trying to replicate something else that we had built. We have extremely competent people. They can do this work. There's this idea that a government employee can't be an advanced analyst, and that's ridiculous. You get yourself five strong analysts, give them just a little bit of training, and you can unleash on any city, that cities are a gold mine of data. They're a gold mine of opportunity, because there is so much that has not been touched for years. Five really strong analysts can change an entire city's world in under a year. Just let them loose. **Liz Crowe:** We've proven that here! **Steve Goldsmith:** Yeah. And Liz, Cleveland came from a place maybe a century behind on this stuff. I mean, you've had a lot to catch up on and done a really interesting job. Just in conclusion, we need strong, talented people like the two of you. They need to inspire change and an appetite for change and willingness to take risks from those in the agencies. They need to collaborate with the agencies and buy them lots of cups of coffee. They need to have a strong champion in the CFO, city managers, mayor's office, and they need to be open to new technologies. What did I miss on that list? **Liz Crowe:** I think they have to have a little bit of fun with it. **Steve Goldsmith:** Ah. **Liz Crowe:** This is cool stuff. To climb in a snowplow is super cool. It's just cool. I know, Brita, you probably don't get the snowplows the way that we do. **Brita Andercheck:** No, ma'am. No snowplows here in Dallas. **Liz Crowe:** To understand the tech that goes into a police car or a fire truck, I look at people all the time. I go, "Do you know how much tech is on a trash truck?" I mean, people are marveled that trash cans have RFID chips in them now, and we can understand routing, and we can understand how often your trash is put out, just the amount of tech that's out there. And if you go, "Gosh, that sounds cool," and you can have some fun with it, you will genuinely change how your city works and how your neighbors experience their day. **Brita Andercheck:** And can I just add to the fun? Because I think Liz has nailed this. You have to have fun. If I were to show up to my job every day and look at the immense amount of problems in front of me, I could easily become overwhelmed, and I could think, "Oh my gosh, why should I even start? This is such an uphill battle." But you've got to think of each of these things as little puzzles that you're trying to solve. And you remember that the people you're doing it for are the snowplow drivers and the 311 agents and the residents and that missing kid that's going to be found, because you wrote a piece of code for a cop. Right? And when you remember that this is fun and you're solving puzzles and that the outcomes are the whole purpose, I think it's really easy to stay engaged and inspired and to show up. I mean, I have the greatest job in the world. I'm sure Liz would agree with me, and it's a privilege to show up each morning and do this work. **Steve Goldsmith:** What a great response. Well, this is Steve Goldsmith from the Bloomberg Center for Cities with two of the country's preeminent public servants, Brita Andercheck and Liz Crowe, discussing data and how to have fun with a garbage truck, I think, is where we ended up. So thank you both for your leadership, and we'll look forward to staying engaged with you. **Brita Andercheck:** Thank you, Steve. **Liz Crowe:** Thanks, Steve. ## About the Author ### Betsy Gardner ![Headshot of Betsy Gardner](/sites/g/files/omnuum10826/files/styles/hwp_1_1__100x100_scale/public/2025-05/Betsy%20Headshot%20resize.jpg?itok=k2OsSp1g) Betsy Gardner is the editor of Data-Smart City Solutions and the producer of the Data-Smart City Pod. Prior to this, Betsy worked in a variety of roles in higher education, focusing on deconstructing racial and gender inequality through research, writing, and facilitation. She also researched government spending and transparency at the Lincoln Institute of Land Policy. Betsy holds a master’s degree in Urban and Regional Policy from Northeastern University, a bachelor’s degree in Art History from Boston University, and a graduate certificate in Digital Storytelling from the Harvard Extension School. See also:- [ 311 ](/topics/311) - [ City Administration ](/topics/city-administration) - [ Civic Data ](/topics/civic-data) - [ Innovation ](/topics/innovation) - [ Operations ](/topics/operations) Share on:- [ Facebook ](#) - [ Twitter ](#) - [ Linkedin ](#)