‘We're all just trying to help,’ local Service to the Citizen winner says

The Unified Government of Wyandotte County and Kansas City, Kansas, like many local governments of its size, is experiencing explosive population growth as well as growth in the number of queries from its residents.

The government’s employees handle around 310,000 calls each year, but many of those routine requests can take them away from their work and increase pressure on their time.

Given that, Crystal Sprague, the unified government’s director of performance and innovation, helped lead the launch of the Government Experience Agent, a digital assistant powered by artificial intelligence that provides multilingual around-the-clock support and can answer a variety of routine questions.

The assistant was first piloted in the government’s Treasury Department, and led to a 30% reduction in call volume, meaning staff had time back for more meaningful work. It will now be scaled across other agencies.

The effort led to Sprague being honored as a winner of this year’s Service to the Citizen award, in the category, “Using innovation technologies to deliver simple, seamless and secure services.” Route Fifty caught up with Sprague to discuss her efforts in government and to hear more about the award.

This interview has been edited for brevity and clarity. 

ROUTE FIFTY: What does it mean to be director of innovation and performance?

CRYSTAL SPRAGUE: Well, for local government, it definitely is a role that we don't see very often across organizations. We see things like a Performance and Management Office if the local government is lucky. Sometimes we just see a smattering of people that have some project management experience. But the organization here made a commitment about 10 years ago to really start, from the ground up, to build what it looks like to do mature project management and technology deployment that also has a strong touch of customer-facing or customer experience knowledge. It's been pretty organic over the last few years.

This is not my background. My background is in criminal justice administration. I actually worked in the probation and parole field for the better part of my career, and then in municipal courts. It's not my career field; however, I think I was selected for this role because not only do I look at technology as a tool, sometimes a toy, but I also have that background of what it means to have a customer experience [focus], how to take care of customers, and how to also think through business processes and change management. I think that one's a big one.

The department grew slowly. It's really starting to get its roots now. It's kind of a little Kevin Costner, "If you build it, they will come." For a long time, we had a hard time with departments being willing to have us come in the door and do things like apply technology with maybe a broader view than the department view, but also to apply things like business process analysis, and utilizing different technologies that maybe they weren't aware of, change management, and things like that.

I only have about seven full-time employees, and half of them manage our Workday product. But we all pitch in and do that project management. We are transitioning from what it means to have an enterprise-wide enterprise resource planning system to what it means to have enterprise relationships with vendors in different ways, rather than having specific platforms purchased by a specific department and then it doesn't cross over amongst departments or functions.

I also take care of our 311 department, so it gives me a little different vantage point of what it means to be customer-facing and how new products deployed to citizens without thoughtful change management can have an impact, because I hear it when they call.

ROUTE FIFTY: I'd love to hear more about this background in criminal justice. How did it prepare you for this very different role?

SPRAGUE: A lot of people just go, "What? You worked in criminal justice, and now you talk about integrations? How is that possible?" But I think it prepared me for this type of role because we're founded in probation practices with asking people to change. A different kind of change — you're asking people to change in a criminal justice environment in a much different way — and you're charged with asking them to change generational behaviors.

But people change in the same pattern, so they accept the change or they defend the change. A change management education, I've found, has really been helpful in this type of role. I would also add that good probation departments do a lot of motivational interviewing training. To be certified in motivational interviewing and behavior modification really brings a lot of being able to get at the crux, or the source, or the diagnosis of the problem, even if it's a technology problem, and you're trying to decide what is not mapped correctly on an integration, for example.

Oftentimes, the people that are close to it maybe need some directed questions and hard questions to get at the idea: is the technology broken, and/or is the process broken, and/or are we holding onto something that we just don't want to let go of yet and we're scared to take that leap into whatever the next thing is? I think those — motivational interviewing, behavior modification, understanding the curve of change — all those have been really helpful. Nobody taught me in criminal justice what an integration was or how to handle a single sign-on login. I feel my way through that with good resources around me.

ROUTE FIFTY: At least you can herd cats, right?

SPRAGUE: Yes! There is that. It's a fast-paced environment, too. Corrections is fast paced. Things change every second of the day, and being able to feel comfortable in that fast-paced environment helps me with project management because that definitely changes.

I will tell you one of the biggest transitions that I had coming from a criminal justice system into the wild, wild world of technology is the rules in a criminal justice system about what happens, when it happens, and how it happens are historic. There are different rules [for technology]. That was a big adjustment for me, to learn that I'm comfortable in the rules, and how do you apply that. That's where I think project management came in for me because there were the rules. I don't have to say them exactly the same way that project management professionals say them or that PMO offices talk about them, but I can be that translation in between, "Here are the rules, and here's my user," and how do I translate in between to get the work done.

ROUTE FIFTY: In terms of the government itself, and some of the things that you're thinking about every day, what are some of the challenges that are keeping you up most at night?

SPRAGUE: Finance is always at the top of that list. Do we have enough money? Are the taxpayers interested in paying for these types of innovations, and why would it be important to them? Financial stewardship always is at the forefront of my mind. Just because I think that there's a fun new product doesn't necessarily mean that it's the right fit for us, or it's the right use of government dollars, or that it fits in the prioritization of the entire organization. Buy a new whatever and install it and install it well, or pave some streets where people really need those streets paved. Those kinds of dichotomies keep me up.

Any kind of technology that is purchased for technology's sake, it's a shiny new penny, and I saw it and I think it's really cool, but it's purchased without a global view of what the government is, and it's not leveraged in that global view. It's purchased with a very slim use case, but it's very shiny. You think it's going to do all these magical things, but an implementation of technology is only as good as the people on the ground that are implementing it.

When I was coming up early, when we bought a piece of technology, it was this thing that we bought, and then we hired the one person that knew how to write SQL reports and they sat in the back someplace, and when it was broken, we just said it was broken and to fix it. Then when it was old, we just got rid of it and got another one. But it wasn't the same kind of optimization curve that I expect and that I think taxpayers should expect.

I often talk to my team about comparison of infrastructure, because I consider big enterprise systems infrastructure. It's technology infrastructure, fiber, whatever. It's sort of like public works — public works has infrastructure.

They have a bridge, and they invest all this money to build this amazing bridge. But because someone over here says they have a specific use case to use their paddle boat to go across the river, then they stop using the bridge. But we've never really asked questions about why are you using the paddle boat, is it a problem we can solve, are you afraid to cross that big line on the bridge, can I help you understand it? We get diverted off into these environments of exception, and so we have this multi-million-dollar bridge that nobody is using because there are now 500 exceptions to go around it.

I talk to my staff a lot about, it's not just the bridge. It's how people are going to use it, understanding all the exceptions that people might have, understanding the fear that they have, because our job is to get the most out of that bridge year after year, which includes maintenance, which includes optimization, which includes problem solving. If we have it as a two-lane and the paint in the middle is yellow and people keep crashing on the bridge, well, maybe we need to look at the paint. I know that puts it down to kind of an elementary way to look at it, but that's really what I think about. If we're going to spend a million dollars on a piece of equipment, or a platform, or a technology, how are we going to really use the investment?

ROUTE FIFTY: I hear what you said about AI and needing the global view. How do you approach AI? How do you think about it, particularly when it comes to government operations? What do you want to see from it?

SPRAGUE: I definitely want to see it deployed in the right places at the right time. It's not a shiny penny. You just can't go grab it off the street and it's going to be fantastic. I want to see it be able to get people back to doing the jobs that they are trained to do. If the Treasury staff is pulled off the front lines of the counter to work pretty serious problems, to answer the phone over and over and over again — it was the manager's choice, that's what had to happen, there was an expectation of customer service. But can I deploy AI safely and effectively to get those people who did not interview at the Treasury's office to answer, "We're open 8 a.m. to 5 p.m.," all day long? Can I get those people back to the thing that they're passionate about doing, because that's where that real strong person-to-person approach is going to be effective. And can I use AI to reliably and safely answer basic questions, do basic functions?

That’s how I've approached it up to this point. I do not agree or feel comfortable with the approach of, "Well, we just like AI, so we should do it." Why do you want to do it? What are your checks and balances? How carefully are you approaching it? Do people even understand what it is? I had an opportunity to talk to the Mayor's office and the Commission a couple of weeks ago, and we were talking about the difference between agentic AI versus other types of AI, and helping them understand that it's really cool that you can make a flyer in Canva using AI, but we also have a responsibility to make sure that it's [Americans with Disabilities Act] compliant, that it checks all the same quality standards if we would have done it internally, and that we think about what information we're putting out there that could be consumed by other models.

That’s the approach that I take. And smooth and steady, not to rush anybody. People are scared. Some people are really, really frightened. How can we talk about that, how can we test it, just like we do any other technology? How can we have a plan to implement it, how can we be steady and smooth and careful about it? It does feel like there are two camps: it's "Heck no, it's going to take my job, I never want anybody to use it in the government, it's going to ruin our lives," and "AI is cool because look at this flyer I made, so we should do it everywhere." How do I find that safe spot in the middle?

ROUTE FIFTY: You mentioned earlier about customer experience. I'm curious for your thoughts on how resident expectations of government have evolved — what people want from their government — and then also how you deal with that as the government. How do you live up to those expectations, or exceed them?

SPRAGUE: Let me put on my 311 hat. I think there is an expectation that government should be as transparent, forward-thinking and advanced as Amazon. If I expect storm debris to be picked up in front of my house, I want to be able to see the truck on the map coming toward my house. There is a disconnect between "I want and I should" and "What I want, and I should pay for."

My approach to try to bridge that disconnect is to consistently educate on hard decisions that have to be made around budget. So, could we have this? We could have this, but how many full-time employees does Amazon employ to be able to do that function, to see the little truck? And what are your priorities? Are your priorities making sure that the jail is fully staffed, or are your priorities being able to see where the public works team is when they're filling potholes?

I always tell people that I'm talking to that it's not for me to decide what you want. You tell me what is important to you, and that's the road we'll go down, and what you're willing to sacrifice or pay for. Now, that's a pretty academic conversation to have when you have two levels of budgetary requirements, elected officials and all the things in between you and the citizen. But I think that if you can give them a common understanding of what your limitations are, and then see if there are things that you can resolve after you have asked good questions to get to the actual problem, then people are generally understanding about your situation, or about the situation that they want resolved.

ROUTE FIFTY: How do you measure success?

SPRAGUE: I do a lot of looking at where we're at and setting a reasonable goal of where we want to go that's meaningful. Again, I'm going to pull back to my probation experience. We used to have this terminology as we were coming up. The majority of probation work is, we used to count how many times we had contact with the offender. But did it make a difference how many times you looked at him? Or does it make a difference in the quality of your interaction, and then what's the true outcome? Great, you saw him three times this week, but what's his likelihood to be revoked or what's his likelihood to reoffend?

When I'm doing things like doing performance measures for 311, what's that caller's likelihood to call back? Great, you answered 70,000 phone calls that year. Great work, you processed a lot of phone calls. But what was the quality of the phone call that you processed? What was the quality of the interaction? And did they call back for the same thing? So that's one way I look at it. That's harder to attain.

Another attainment is, for example, in the Treasury project: can we pull down those telephone calls by a certain percentage, freeing up that magic number of a full-time employee to get them back on the front window, or to allow those conversations with constituents or callers to last 10 minutes instead of two minutes? Two minutes is not even long enough to say your name, and that's what our call takers were averaging. They could only be on the call for two minutes at max, if they worked every day and never had a holiday or vacation. That's not a quality interaction with citizens. You can't really ask them even four questions to get at the root of the problem.

ROUTE FIFTY: You've won the Service to Citizen Award. How does it feel? How do you reflect on this honor?

SPRAGUE: I'm very lucky and privileged to be exposed to a group of people at the federal level that recognize that the problems that they have are the same problems that I have; it's just a magnitude of scale. I'm very honored, and I have felt very privileged to be in such an amazing group of folks that have won these awards.

It's a reminder for me that it doesn't matter if I'm working through the Social Security Administration or I'm trying to call my local government to find out who's going to pick up the trash alongside the road, we all serve citizens that are trying to get a task completed. It doesn't matter if it's a task of paperwork or a task of calling and getting it completed. We're all just trying to help.

It's a good reminder for me that the federal government isn't this massive being that doesn't have the same problems that we have. Same problems, just different magnitude, and the solutions a lot of times are the same. It just might be a different amount of money, different scale of technology implementation, but at the end of the day, it's still the same approach.

ROUTE FIFTY: What's next on the horizon for you? What are you thinking about, what are you worrying about, what are you excited about?

SPRAGUE: I'm really excited about building an enterprise relationship with a vendor that we use named Granicus. I think this is a new horizon for them, and it's a new horizon for us. But as we've acquired different products from this one vendor, and they're all based around communication and interaction with citizens, I happen to be more excited and passionate about those types of communication tools than maybe, you know, another type of technology.

If you look at an enterprise relationship with a company that spans all these different products, and then you're able to put an agent — a controlled, safe agent — on top of all of them, how it will plant the seeds of change in how service can be provided with low budget at a local government level in a way that I don't even think that I imagined.

I'm very excited about this. And even though I am close to having my 10-year in for retirement, I hope that the seeds that we're planting today — be it with that particular vendor I mentioned, or just planting the seeds of how to look at your interaction with citizens, whether it be an aged citizen that's interacting with our Aging Department, or somebody that's calling the Mayor's office, and being able to deliver unique and personalized service to that person all the way across, and then getting them to resources that they wouldn't normally have been exposed to just based on a singular interaction — I think that's going to be amazing.

We need to be ready. A lot of what I'm doing today is not only exploring what that looks like, but getting the organization into a contemplation or pre-contemplation stage in change, where they're starting to get ready to accept whatever the next thing is.