How AI Powers State and Local Government Innovation

Hey, happy Friday everyone. This is Joe Morris. Dustin is on a well-deserved vacation today, but we're here to talk about AI, and I don't think there's a hotter topic than AI right now in the public sector. From chatbots to streaming suggestions, AI is playing a big part in our lives. It's also helping governments run more cost-effectively and efficiently. Joining me today to talk about AI is Kim Majerus, the AWS VP for State and Local Government and Global Education. She'll be sharing with us how the work world's most comprehensive and broadly adopted cloud, is supporting government modernization with AI. So thanks for joining us.

Well, thanks for having me, and what a better way to end the week and start a long weekend. So I appreciate you inviting us.

Absolutely. So maybe you could spend a little bit of time just sharing with the audience, tell me a little bit more about yourself and your role.

Sure. As you mentioned, I lead the US state, local, and education business. I also have responsibility for global education. But most importantly, I lead an organization that accelerates innovation learning environments, research, academic medical centers, and a variety of citizen services. My organization is quite diverse between higher education, K-12, edtechs and govtechs, independent software vendors. The way I look at it, my team focuses in a couple of areas. One, for the actual institutional side of the business, but then too, any of the commercial organizations that are laser-focused on supporting our strategy to help support governments and education institutions around the US and education globally.

Well, can you give us a little bit of an overview in terms of what you're doing today to collaborate with those governments?

Oh gosh, as a company, and for those of you, many of you know, AWS was launched in 2006. we were actually an output of Amazon.com. But when we think about the public sector environment, we actually launched in 2010, and we primarily focused more exclusively in the US federal government space. In 2017, I joined AWS to start up their SLG business. Today, we now support over 7,500 government agencies, over 14,000 academic institutions, and over 35,000 nonprofit organizations across the world who use AWS. And that's in over 200 countries and territories. You know, when I think about governments and the citizens that they serve, you know, they should have the same great infrastructure services that are driving innovation inside of public sector and even outside. So when we think about our work with governments, it's really to work with them to continually modernize the way that they conduct business. So we're supporting modernization by providing the most innovative, efficient, and most effective solutions. You know, first and foremost, citizen data and security is always top of mind for AWS. And for me, my customers, it's the number one thing that I keep my eyes on. But we never compromise on privacy and security. Likewise, we never compromise innovation. It's that relentless innovation that requires just that focus beyond product development. In today's rapid-changing environment, we have to make sure our customers really adopt that builder's mindset. As we know, our public sector leaders, they are focused on solving challenges. They are the professors and the opportunities that they have for our citizens. Really, that's what we get to work on. They give us a problem set, we work with them to help them support.

That's great. I think of the work that I know you've worked with eRepublic on and GovTechCon is focused on those priorities, right? We look at our CIO priority list, you see cybersecurity like you mentioned, you see automation, and you've certainly seen cloud there now for the better part of a decade or more on that priority list. What are some of the unique advantages that cloud's offering to public sector entities right now?

Yeah, you know what, I already mentioned the security aspect. But when we start to think about those mission systems, you know, cloud provides them a more responsive environment. It allows agencies to actually move and iterate so much more quicker than they have in their legacy environments. It's responsive to new policies. It's also proactive against cyber threats. And, you know, as a result of all of that, it does allow the agencies to actually reduce costs. significantly, and it provides the best tech without a massive, long, lengthy procurement process. And I'd have to say, if you think about what we experienced during the pandemic, that was a perfect example of how cloud supported states who already started their journey to AWS. Those states who had to respond quickly and differently were able to do that because they had the agility of the cloud. Whether it was Health and Human Services— to support citizens, or think about the call centers. When people lost their roles or they couldn't go back to work, think about the amount of call agents across every single state that had to quickly traverse from being in a call center to working from home to support overnight. They had to do that literally in order to support citizens. But I mean, that's the exciting stuff that we were able to help support. Because that's really what made a difference during a very difficult time. And I'd have to say the leadership across government, whether it's state or local, the state or local organizations, they really moved quickly and they innovated. They had to because so many lives and so many people's livelihoods depended on it.

You're right. I think having the right foundation in place was absolutely a prerequisite to success throughout the pandemic, and it continues to be. Let's pivot a little bit to topic I think that everyone is focused on right now, and that's artificial intelligence, generative AI, and all the things that you can't seem to escape in terms of the media and in our personal conversations. And maybe you can break it down for our audience in simple terms. How can the workings of AI benefit government operations?

Well, I think the important part is AI and machine learning have really been a focus from Amazon for over 20 years. And I think it's hard for people to kind of correlate the two, but let me, let me try to give it to you in a different way. So when you think about AI and machine learning, when we think about our retail environment, it's really how it was all founded for us and it findings that allowed us to share innovation with all of our customers. So when you think about the algorithms in your shopping environment to which Amazon Retail operates, Now start to think about how that could actually change the environment to which our citizens and our students are trying to evolve. When we think about what AI and ML actually is, machine learning is the science of developing algorithms for statistical models that use computers to perform these tasks without any explicit instructions, but it really relies on patterns and inferences instead. This allows them to predict better outcomes and more accurately given into an input dataset. When you think about the deep learning, that's really a subset of machine learning and it's modeled more on the human brain. Deep learning algorithms analyze data with a logic structure that's simple to use by humans. When you think about intelligent systems or what they call neural networks, it helps provide a process of information in different layers. But hey, as technology continues to evolve, so do the capabilities of AI/ML. But that's where we've invested heavily in the development and deployment of AI machine learning for those decades in those customer-facing and quite frankly, our internal operations. From the search engines when you shop on amazon.com or the AI-powered robots that we actually use and optimize to fulfill in our warehouses, it's the opportunity that we're starting to see where government agencies can improve their services. Think about a time instead of when people go to a website and say, what services do I qualify for? What if it changes to say, hey, these are the services that you qualify based on life events or opportunities that they've recognized in the services that you've been pulling off from the past. I think the important part to note, and I think states and counties definitely see this, the services that they were supplying over the decades, may not necessarily be the services of the future. And think about it from an education perspective. You know, if we could use this data to help hone in on providing a better student outcome because we understand people learn differently, think about the power of what we're able to put a more skilled workforce out into the market as we start to customize and personalize their learning experiences.

I love all that. And as someone who's, you know, looked at the power of data and its implications in government to the programs that you mentioned, the educational outcomes, I think it's hard to argue with. But I know what our audience really gravitates to are those specific examples. Do you have any specific examples where states or localities are harnessing the power of AWI through the work that you're doing?

Oh gosh, I have several. So if you want me to stop after the first 2, I'm more than happy to. But, you know, we've actually helped hundreds of thousands of customers of all sizes across all industries. to really be able to use ML and AI with these leading opportunities. Let me give you a couple of examples. Let's think about first responders who are like the hardest working public servants in our country. They head towards emergencies when others are running away, and often they're the first contact anyone ever has by calling them on 911. The International Academies of Emergency Dispatch Reports that some of these call centers are struggling with as much as 30 to 50% vacancies. These are people that we need on that front line. But as a result of these challenges, we are helping customers use Amazon Connect. It's an— you know, I talked about how we were able to help agents go from call centers to working from home during COVID Well, they're also using Connect in these environments. So It's an AWS service with AI and machine learning solutions that reduce the burden of the 911 call center operators because they're actually diverting and handling some of those non-emergency calls. You know what? We have data that shows it's working. We have a community that is— I mean, if you'd like to call them, we'd love to have our customers talk on our behalf, but they actually saw a 10 to 12% reduction in daily call volume because of non-emergency calls being offset by the solution. You know, in an emergency, time matters, and even a few seconds or minutes can mean life or death. So being able to equip the organization that are taking those first inbound calls and diverting them to solve for a different challenge through the capabilities of machine learning and AI responses that have already been logged because they're simple tasks, That makes a big difference. But when you think about it that way, think about it from a public health and healthcare perspective. So we all know Medicaid is unwinding and it's a national effort to actually recertify every single eligible recipient of Medicaid. I mean, we're talking nearly 90 million people. That means forms, mailings, calls, a massive amount of data that has to be processed. So when you think about the scale of unwinding efforts and automation and the scalable digital engagements, that's, you know, like, it's key because if you are unwinding these efforts, you could be unwinding people that are genuinely in need of those services. So with an automated outreach to the members that are currently getting Medicare, Medicaid, it doesn't have to be in form of a call or even a mailing. we're actually using new ways to engage those Medicaid recipients. So here's an example. Martha, my friend, she's a Medicaid beneficiary. She moved in with her parents during the pandemic. If the state is trying to get ahold of her solely through physical mail with an address that they have on file, Martha loses coverage because it's time sensitive. So now if the Medicaid agency sends a letter, They only have what they had on file, her old address. She's going to miss that opportunity to, to actually recertify. So with Amazon Pinpoint, which is a multi-channel communications tool that helps in this scenario, it uses automated texts and emails to help customers perform outreach. It's collecting address information to complete all Medicaid renewal requirements. But Amazon Pinpoint and agencies using it really saw an opportunity to help streamline their engagement with their recipients and actually retaining coverage. That is so important, especially we know those that were receiving Medicaid tend to require Medicaid as they move forward. So that's just, you know, one example. Last one I would say is John. John is a Medicaid contact center agent. You know, he uses Connect. He fields calls from members and assists them through the renewal process. He loves his job because he's helping those recipients. However, John is spending majority of his time answering some of those same simple and common questions. He spends less time to make sure that our people, you know, people that are needed are actually being covered. So all the automation, contact centers have long wait times, increasing backlog of these questions. With Connect, Amazon Kendra, Amazon Lex, it all helped them really refocus and take those language processing skills to be able to produce a highly effective triage of the calls that can be answered differently. It helped his workload. And at the end of the day, the people that actually needed to retain coverage, they were able to retain it. That for me is time and efficiency of those that are focused on mission to take care of their citizens. So those are just a couple of examples. We have so many more to come back with you, Ed, and you could always find them on our website too. But like I said, it's extremely important for us to have our customers talk about the services and capabilities that Amazon has been able to supply them.

Well, I appreciate you sharing those ones, and we'll make sure to provide a link in the show notes where our audience can find the other examples that you referenced on your website. The scale of some of the issues that you just talked about though is impressive. And you also referenced a community. Do you collaborate with other technology partners, vendors when it comes to providing these comprehensive AI solutions to your government customers?

You know, absolutely. And in specifically in the state and local space, you know, outside of us being 100% customer obsessed, We want to make sure what's working best for the customer is also engaging with the partners that our customers look to support, right? So a lot of our states or cities and counties, they tend to be buyers versus builders. So what we do is we partner our customers, these institutions, up with over 100,000 different partners that can serve their needs. from across the globe. Let me give you an example. When I think about— there's a partner called Recore Systems. It's the global AI technology. I'm going to focus more on the AI stuff since we were talking about it. They're a public sector partner and their mission is to provide insights that build safer, smarter, and more effective cities through their intelligent infrastructure. In 2022, I'll give a call out to Philadelphia Industrial Development Corporation. They started a pilot where they were trying to analyze traffic patterns. Now, for all of us, and I know you're in California, so I can only imagine the traffic challenges that you see. But in Philly, they were trying to understand the travel patterns at the Navy Yard. They wanted to understand the volume, the flow, how many large vehicles, how many small vehicles. They want to gain visibility into the impact of the road closures that they were planning. And, you know, in the East, We tend to deal with a lot of different climate challenges with snow and rain and what have you. But I think the important part to note is when you start to work with a partner, they have done this in other areas and they want to be able to show you what they could do differently. So they were able to work with Recore and across the Navy Yard, Recore put smart devices and leveraged AI to collect and de-identify data, mind you, to analyze it. The data transformed the critical traffic insights that they were able to see real time. So it included how many visitors, how much traffic volume, what type of vehicles, how many electric vehicle volume, how much greenhouse gas emissions. But at the end of the day, the end customer used this data so that they were able to plan better, budget better, and it actually allowed them to understand where they needed to change their service. For example, do we have enough electric charging stations, or do they have to install more safety systems? And I would say, if my recollection is correct, in February 2023, that's this past year, they had 25 smart devices installed, and they are collecting it and feeding it through the AWS Cloud. Ricoh then leverages those services and provides better insights. Now, when you think about it from a roadway perspective, especially you in California, Understanding what challenges you might be uncovering as you're driving from point A to point B is extremely important, but making sure the cities and the counties are serving them better is really us as citizens trying to make sure we're optimizing our own time.

Earlier you mentioned this customer focus. You said you're laser-focused on their needs. In the space of AI, What's setting you apart from the other cloud service providers or other providers in terms of, you know, helping government accomplish their mission?

Yeah, it's a great question. We have a very unique way of engaging our customers, and what we call it is the working backwards process, because a lot of times people don't necessarily know what problem they're actually solving for. So we genuinely show a deep interest to understanding the problem statement, Because it's not about what I have to push you, it's really about what is the right thing that you need to solve your challenge. When I think about our deep listening skills, making sure that we fully understand the challenge that they face, then we pull in our solution architects, our government advisors, and our experts to work to find the solution that's best to solve the problem. Now, it's the same principle and it's very helpful when you're considering AI and ML. You know, the first step is identifying what the problem is, and once you've determined that, then you work backwards to the solution. And really, sometimes it'll be AI, sometimes it won't be. You know, when I think about it, AI is not a fix-it-for-everything, but it's an amazingly powerful tool when applied in the right situation. You know, after a solution is found, it usually needs to be refined, and at AWS, we refine these ideas we create a proof of concept, we test it, we refine it again until we're sure and the customer's sure it's the right solution.

But you know what?

Working with AI is very similar. When you have a solution that involves AI, it needs to be tested and refined until you're comfortable that it's working. For me, that's extremely important because when you think of a problem statement, it's not just throwing something at it. It's really about making sure it's the right way to use it. I think that's the part about how we are hearing all these conversations about AI and ML. How do we use it? How is it best? How is it secure? How are you ensuring that if you're running these models, you're running them in a secure environment that never leaves your environment? Because we know our state and local customers have to really focus on ensuring secure, the security and the privacy of the citizens that they serve.

So before we wrap up on this Friday afternoon before a long weekend, is there anything else that you'd like to share in terms of AWS's vision for the future of AI in the public sector?

Oh gosh, there's so much, but you know what, because I know we're short on time, I wanna leave you with 2 things. Now, if you're out there, I'd love for you to go to LinkedIn. Look me up. I literally this morning posted a link for an event that we're holding specifically on generative AI. It's on June 6th. It's an event and I'm confident if you go there, log in, take the time to invest in understanding how AI can help you and your problem sets, that would be number one. Then the second one, and since you're out in the Sacramento area, I'd love to invite anybody out there to join us For Imagine, it'll be July 11th through 12th in Sacramento. It's a free event, no cost, 2 days where you're going to get experts to talk about all aspects of how to solve for your customer challenges. We will have ISVs, we will have keynote speakers, and we'll have our customers talk about how they're using services to better serve their citizens and their students across the US. And actually, Very cool.

Is there anywhere else that our audience should go to learn more?

You know what, we're gonna drop in, there you go, there's a link, go.aws/3bnzq. So if you have a chance, click there, we'd love to hear from you. And I look at it this way, if you have a problem statement that you need solved, reach out to us. We love solving problems. solving the biggest problems because that makes the biggest impact. And we want to make you be successful because at the end of the day, we're all citizens of this great United States. And I want to make sure that every single public servant serves their citizens in the best way that they can.

Well, Kim, that was great and a wonderful message to kind of conclude our episode today. So thank you very much for joining.

Thank you for having me.

Have a nice long weekend, everyone.

And you.