Podcast Interview with Duane Powell, CIO of City of Hiliard

Transcript:

Dan Greenleaf: Hello, everybody, and welcome to AI  Rising. I'm Dan Greenleaf, Co-founder and CEO of AI Rising.  

With me today, I have an exciting guest, Duane Powell, the CIO of Hilliard. We're going to talk a little bit about an exciting project they have called the AI Sandbox, but before we dive in, Duane, do you want to introduce yourself?

Tell us a little bit about your role at Hilliard and what you do. 

Duane Powell: Yeah, I've been with the city of Hilliard for coming up on six years now. 

I work with a great group of people across the city. My primary responsibility is IT support and supporting all the infrastructure in the digital space across the city to provide services,  and I'm also very fortunate to get to participate in our city initiative of Hilliard City Lab. 

And working with companies with some really great ideas, growing them in the city and becoming a part of our ecosystem.  

Dan Greenleaf: Sounds really neat. And I'm always excited to talk to folks in government. You know, those of us in the private sector, we're always looking for the opportunity to make more money. But when you're in the public sector, it's about making the world better and about improving our community. 

Appreciate you taking the time today, Duane. I was excited back at our conference last year to have you come talk about this AI sandbox you built up in Hilliard and how it's allowing startups to really get a first access to AI tools at an affordable price. Can you tell me a little bit about what the AI Sandbox is, just at a basic level. 

Duane Powell: Sure. So, you know, the first thing to say about the AI Sandbox is not a replacement for traditional cloud providers. But it is a test environment where we can do proof of concept around AI technology. And we really had this idea. It came about from working with some of our customers that are in our ecosystem of Hilliard City Lab, who said, one of the bigger problems that we have in moving quickly in the AI space is this risk of a cost burden when we're in the traditional cloud environment.  

So this is prototype code  and we're developing it and we have to work very methodically about how we do that. Because if we make a mistake and a process runs away from us, what we would think would be a traditional build can be astronomical, and unfortunately, you know, put some of these smaller companies out of business.

So our goal was to create an environment where we could do this proof of concept and let them test, let them work out the kinks , and the only penalty is maybe a phone call saying, Hey, you got a runaway process here. Maybe, maybe turn that off.  

Dan Greenleaf: That's really neat. So what exactly services are included in the sandbox? What can you guys support today?  

Duane Powell: Well, today we're working on just traditional compute, right? We don't offer any extra services around that. You build your environment within the containerization that we have and can run however you want to go about doing your product testing.

One of the services that we're looking to add this year is again, in conversations with a lot of our customers within Hilliard City Lab was,  ‘Hey, it's great, but how do we understand what our cost is once we go to productize this,’ and that is important, right?  ’How much compute will it cost when I go to a traditional cloud provider?’ So that they actually know how to price that.

So that's one of the big services I think we're going to be adding this year. That can add a lot of potential to that environment and be very helpful for our partners in developing on this platform.  

Dan Greenleaf: Yeah, that's pretty amazing. I know a lot of people don't realize the actual computing cost to run these AI models, these large language models that just take enormous amounts of compute to produce a very quick answer. 

So tell me a little bit about what technologies folks can use or what exactly they can do on these compute boxes you provide.  

Duane Powell: Well, right now the vision around this is within Docker containers. So we're going to containerize it per company. And we've been testing that with a few companies to date.

One of them, Ubivision,  which specializes in AI video. In identifying, say, anomalies within a video stream  or patterns or in the retail space, being able to identify certain customers  or products or how long somebody waited in line,  how long they were with a customer service rep, those types of things, those statistics that a lot of our retail partners are very interested in on their customer experience. Not my place to divulge, but some very big companies are working to be able to do that. 

Dan Greenleaf: That's really neat. So I think you mentioned a specific use case for public safety around the Ubivision work you're doing.

Could you tell the audience a little bit about that? 

Duane Powell: Sure. So the city of Hilliard has well over 400 cameras across the city. And traditionally we use those cameras are very reactive. In cases of, you know, there was a traffic accident. So the police can pull that video and see exactly what happened.

 Maybe there was some vandalism across the city or something like that. And we can identify what happened after the fact. But it's not very proactive. It's not very responsive to when the event occurred. This really came about as a result of an unfortunate incident in which a young man hit the curb, going about 120 miles an hour, best we could calculate. 

It launched the car over the roundabout. But unlike television, the car didn't just land, it came apart and caught fire.  Fortunately, the young man was able to get out of the car. But he collapsed. He did survive and that's very good news, but what we noticed was he laid there for over 30 minutes before another car came along. Because this happened in the wee hours of the morning when there was no one around to call 911.  

But our cameras saw all of it from the moment it happened. So we took that challenge to Ubivision, and Eric Wagner's the CEO of Ubivision and said, Hey, how can we use these video feeds in a way that we can provide our services. In a more proactive manner. So we're looking for anomalies.

We're not looking for individual people. We don't do facial recognition. We're not tracking people. We want you to be in the city of Hilliard, but we don't, we don't care. But if, if you're walking along the city street and you fall down, we care about that.  And if you fall down and you don't get up, then we really care about that.

And we need to notify the appropriate person to maybe at least look at that camera and see what's going on. And so we can provide our emergency services in a proactive manner. As opposed to waiting on you to call 911, which takes time, or someone to happenstance along and be able to see that you're in trouble and call 911. 

Dan Greenleaf: It's really amazing that you're actually trying to save lives with AI and doing it in a low-risk format too, because it's one thing when you control someone's pacemaker and you might cause major medical issues if you don't control it right.  It's another thing when you alert folks to a potential issue, and if you fail to alert them, it's no different than what would have happened anyway, so. 

I love these 'high impact when they're successful' and 'very low to minimal to no impact when they fail' opportunities. That's really amazing. So excited to hear about it. Tell us a little bit about if someone wants to get involved and use the AI sandbox, what are the requirements?

What are the costs? How do I get started? 

Duane Powell: The cost is really easy. We don't charge for access to our AI sandbox. We do require a presence in the city of Hilliard. That can be a co-working space— that can be whatever. But Hilliard City Lab is all about growing our ecosystem and about businesses within the city of Hilliard.

So we do require you to be a part of our ecosystem to participate. And then it's just a paragraph.  What do you want to do with it? Not a page, not five pages, not a long, it's one paragraph. What do you want to do? And we can, we can move forward.  

Dan Greenleaf: Wow, you guys try and make it easy. I like that. 

Duane Powell: Fast and easy. 

Dan Greenleaf: Yeah, so what else are you most excited about, you know coming down the pipe here with this AI sandbox? Are there things in the works that you can't tell us about yet? Are there things that you can? What's most exciting here to you?  

Duane Powell: Oh, I think I mean, you know, AI is is pretty exciting in general, but I think what's exciting about Hilliard City Lab and the companies that are coming is is be able to work with all these really great minds that have some just fantastic ideas.

That all they need is a little bit of a leg up to get them started. And, and that's really what we're all about is, is we just want to help a little bit. We, we're a little bit different, right? We don't run an incubator in which you have to use our real estate.  The whole city is a part of Hilliard City Lab. 

So it's not just about the AI sandbox. It's about any of our infrastructure. So. Maybe you have a great idea that is AI, but you need access to someone in logistics or someone in public safety. We have those expertise so we can get you into that ecosystem and to get you talking to the experts and get them evaluating that product right away.

So there's a lot of companies that are working with us right now. We have the bandwidth to work with a lot more and, and we're just really excited about  2025 and everybody that's headed our way.  

Dan Greenleaf: Awesome. Well, excited to see the new things coming out of Central Ohio and the Midwest.

Really has a lot to offer in the world and can help build amazing solutions. So you mentioned a few other things besides the AI sandbox. Is there anything else from a Hilliard perspective you'd love to promote while we've got you on the channel? 

 Duane Powell: One of the things we're really excited about, it's kind of in conjunction with  Ubivision, is the Lighthouse Technologies and their vision controlled airspace for drones.  So it controls the airspace below 400 feet.  So if you think about, flight controls, radars, traditionally are very expensive. And we're able to with this system control the airspace and identify aircraft and other things like birds and other drones that may not be registered in our airspace. 

And we're really hoping that this year we can get the authorization from the FAA to fly beyond visual line of sight. So if you think about that, you know, cameras that we traditionally have today are fixed. But once something is identified, if we could get a drone on site to help coordinate those activities and to help our public safety respond, not only to an emergency, but just to manage the vast infrastructure that encompasses the city of Hilliard.

So we're really excited about that.  

Dan Greenleaf: Wow. Yeah, I can only imagine if, you know, there's a major accident, multi vehicle, you can have a drone in the sky before first responders get there and they can know what they're walking into.  It's an amazing idea. I hadn't even thought of that.

Duane Powell: We have three of those towers, in operation today, across the cit,y attached to the city's fiber network that's controlled by a central air command space that they have at Converge Technologies. And we've got our application in, so we're hoping that that gets approved this year. 

Dan Greenleaf: That's really neat. Awesome. Hey, well, I appreciate you taking the time, Duayne, and look forward to seeing you out in the community and all the cool things that you're doing with Hilliard. 

Duane Powell: Well, thanks for having me, Dan.  

Dan Greenleaf: Awesome. Well, you have a wonderful afternoon and we will talk to you again soon.

Duane Powell: Talk to you. Bye. Bye.