Jason Meshell with Microsoft

SUMMARY KEYWORDS

Industrial Talk, Microsoft Azure, cloud adoption, data centralization, AI power, nuclear energy, data centers, IoT data, AI optimization, power requirements, data transformation, cloud bursting, data security, AI applications, user conference

00:00

Welcome to the Industrial Talk podcast with Scott. MacKenzie Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots,

00:21

and let's go right once again. Welcome to Industrial Talk. Thank you very much for joining and thank you once again for your continued support of this platform that celebrates industry professionals. Yes, you are bold, brave, dare greatly, I say it all the time. I am a big fan of who you are. You industrial Pro. You're changing lives and you are changing the world. We are here on site right now. It is accruent insights, which is a user conference for all of their wonderful solutions. And they have customers. They're talking about solving problems, and we are in San Antonio, and it is a great event. All right, he's amazing. He's never been on the podcast before, but he should have been. His name is Jason Michelle, and he is with Microsoft Azure, so let's get cracking. Yep, you're having a good conference. We already had that conversation. Yep,

01:16

great conference.

01:18

I like the people. You know. What I've been noticing about this conference, because everybody's focused on the customer, like, accrued is really customer Central? Yeah, obsessed with the customer, which is a good thing, not not at a negative way. It's just it. That's the sense I get.

01:33

Yeah, carrying, not stalking, right?

01:36

Yeah, there you go. Yeah. Let me write that down. Yeah. Carry, not stalking, that's

01:41

right, obsessed, yeah, making sure that you understand why they're here, why they would spend money, time, energy to get here. They prioritized it. So it's important to prioritize them.

01:52

No, absolutely spot on. Well before we get into the conversation about your relationship with a crew, it give us a little background on who you are sure.

02:00

So been IT professional for 25 years. Spent about a decade at IBM, and then when I started to see the cloud really kind of emerge, I saw Microsoft's strategy around it, decided that's the group that I wanted to be with. So jumped on board. Steve Ballmer was still at Microsoft, and then right after I joined, Sachi took over the helm, and everything was cloud first, mobile first. So everything's been focused that way since I joined, and so I've got to see the evolution truly, of an industrial strength cloud from the ground up. It's been a great journey.

02:36

Help the listener understand what that is, sure. So

02:39

back in my IBM days, we sold mainframes. We took a bunch of data all across companies, all across the world, actually, we put it in one central location called a mainframe. What happened was we went we distributed all that information across client, server world, and you had servers under everybody's desk, servers in closets, and data was spread all over the place once again, which was great. It was, it was an evolution. But what people realized is, what if we could centralize all of that again, in the cloud? So it's literally like, you take a mainframe that's like, physically hosted in an area, you move it to the cloud, which is a cluster of compute and storage across the world that you didn't can make smarter, faster, more secure, and it's like, literally, you take a mainframe, put it into the across the internet, and you have the ability to aggregate, and the economies of scale give you the ability to do things much faster than you would before. So that's really the cloud.

03:40

It seems to me that great, great explanation, what fascinates me is is there's still some reluctant, reluctant reluctance to get into the cloud. Some people are still talking about on prem and making that safe. I help me understand why I should me. Scott MacKenzie, need to get into the cloud.

04:04

Yeah? Easy. When you walk out of here and you drop your phone on the road, when you're getting into your Uber and it runs over your phone, yeah? What? Where are your pictures going to be? Are they all on your phone? No. Where are they at? And they're in the cloud, yeah? So you don't want to lose the pictures of your of your wife, of your grandkids, of your hunting trip, of the painting that your grandpa, you know, painted, but you know it burned up in a fire, but you have a picture of it. The digital representation of your life is stored in the cloud. And so why would you want to put it all on a device that could be damaged, when you could have it put into a space, whether it's going to be protected, and then they replicate that in multiple locations across the world, so that there's a tornado in Kansas and it hits the data center where your pictures are stored. There's another copy in Tennessee or in Utah, so you can still get access to it. You get your new phone, you upload it, there's your contacts, there's your pictures. That's the power of a cloud. On a very economic and consumer scale,

05:04

where is, and I don't even know the answer to this, I just think everything is going to be a server farm. Eventually, there's just going to be building up, up, up the server farm. The demand for power is increasing. Sure. How do we address that? Well,

05:21

just this week, whether you, I don't know if you saw it, Scott or not, would you remember the movie The China Syndrome? Yes, when Three Mile Island melted down. So

05:29

saw the story Microsoft,

05:30

right? Yeah, Bill Gates, Microsoft. What do we do? We're gonna go in. We're gonna take this mothballed nuclear reactor, we're gonna make it safer. We're gonna partner with the company that's creating the nuclear energy but that's going to be specifically to power AI, to help the world transform and change and bring things to market that we never thought possible. So, yeah, that's about, that's what it's for, that's that's what it's for, for AI, for AI,

05:56

the power to you got it needed for AI, you got generative AI, yes, I didn't know that. Yeah, that's what it's for. I just, I just read the headline. Wow, check that out. Well, look at that. What that's distracting? Yeah, I wish I could change the that's incredible. People are on stilts, and it looks pretty doggone, cool, and it's beautiful. Okay, yeah, go. Now I'm all distracted. That's okay. I can get you back your second all of a sudden. I didn't know that that was the

06:21

case. Yeah, it is. Yeah.

06:24

So we have this facility that they're going to address the power issues, but that's still just in one area, just

06:30

07:29

it is amazing. But then we still don't, we haven't addressed the infrastructure as well. Sure, the transmission infrastructure, yep, all of the wires, all of this, that's why you put it close, yeah, that's why you put it close. Yeah. This is cool stuff. Yeah, talk to us about the relationship you have with what. What are we talking about? Well,

07:46

well, so like any relationship, it typically starts small, right? And you start to build trust. And so a couple years ago, Marvin Clark, the CIO and I started on a journey where they were looking at cleaning up their data. They were doing big data projects, they were doing some machine learning. And as they as we started to go through that evolution, they wanted to take some of their older applications and get them up and into the cloud. So we started helping them to transform some of their applications through a DevOps process to the way they were building certain things. We were partnering with them. They have a great team, and they've added some really good new talent. And so as they started to transform that, we were starting to move some of those applications up into the Azure Cloud. And, you know, year and a half ago, this thing called AI comes out of Pandora's box. It's just explodes and it goes everywhere, right? Well, because Marvin did such a good job and his team of cleaning up the data, they were able to take advantage of AI earlier than a lot of other companies, because the data is clean, garbage in, garbage out. If you have bad data, you're gonna have bad results from Ai. They're getting good results out of AI because they cleaned up their data ahead of time. So that's one of the reasons that our partnership, I think, has grown faster than some of my other partnerships, is they did the hard work ahead of time to clean things up anytime you move, whether it's from on prem and into the cloud, or you move from you're trying to move grandma's apartment, and she's taken the entire house down into the condo she bought at the beach. Yeah. What do you do at the piano? Yeah, it doesn't fit, right? So you have to figure out how you're going to move certain objects up into the cloud. So you have to refactor a Martin Marvin and his team did a really good job doing that. Marvin, Christie and the rest of the

09:18

team. Do you see that there's going to be because they've done a good job at that. And it just seems to me that there's going to be a continued demand for your solutions. Yep, continue to expand, continue to because now everything is just data. Yeah, it's just yeah, grab as much data as I possibly can, yep, using use some sort of algorithm or AI or ml to try to analyze that data as much as possible. Yep, spit out results and just keep doing that, day in, day out.

09:53

Yep, improving it. Yeah, constantly. So

09:56

how do you as as Microsoft just take that into consideration, because. There, there's a tsunami of data.

10:02

Oh, without a doubt. And so what, what accruent service channel and Gordian, you know, facilities, asset life cycle group is they're looking at all of the facilities that a company may have, a college, a hospital. They're also looking at all the assets that they may have inside of those buildings. And a lot of those, those, those assets are now instrumented. 20 years ago, they weren't now they are. So there's a data flow, the Internet of Things, the IoT data stream that they're able to capture. They can take that, harness it, put some intelligence around it, put some context around it, and serve up some very interesting things for their customer base. So it's like, oh, this college, they have all these facilities, and they're full nine months out of the year, three months out of year, they're dormant. Well, what do we do with those? How do we help our community? How do we open these spaces up so that we can help with summer schools and different things like that? The Performing Arts nonprofits, they can use these buildings, and if they have the ability to, oh, they're censored, let's let the community know they can use them. Colleges right now are using FAL and accrued software to start to look at these things and go, Oh, we can be about we can be a better citizen to college. Can by making sure that we understand where our needs are and, Oh, guess what? The downtown University is no longer as populated as it once was. We need more space at the suburban university so we can reallocate resources. We can reallocate where people are, where they need things, against all the different assets and the life cycle of all of those facilities. And that's what these guys are doing. The ability to have Cloud gives you the ability to take massive amounts of that data, which you can't put, say, on a server under your desk. You spin it up into the cloud temporarily, you run all your calculations, and then you spin it down. It's like renting an Uber. You don't need to buy a car to go to the airport. You rent the Uber for once, it drops you off the airport. What a crew is going to be able to do is take the data that they need, the need they have for that four hours, put it up into the cloud. Microsoft will process it. They'll have the processing power in the storage. Then when they're done, they shut it down. Uber goes home. They have the ability to do massive transformations without having to pay all the time for servers sitting in their own data center. We'll, we'll handle that part.

12:09

I'm noodling on what you're saying. So I'm sending data up in the cloud. We're processing that data in a way that needs to be processed, making it relevant, making it real, whatever might be tactical insights, and then you say, shut it down. Yeah. What do you mean by that? Well, what

12:25

I mean by that is so storage. Storage is, you know, quasi permanent, if you want it to be, but the actual computing power to do some of the projects that you're looking at take a massive data set. It's like putting 100 PhDs on a problem in a room, but the PhDs are the AIS that are going to be working for. You take all the data from all the buildings, let's say it's a hospital system that's 300 locations. Take all the data, you load it up into the cloud in one data set, and then you put the throughput, and you put the PhDs of the AI around it to start asking you questions. But you don't have to leave all of that data in the cloud forever. You didn't pull it back down, and you don't have to pay for it anymore. So it's bursting. It gives a crew at the ability to have massive power of Microsoft behind them. But they don't have to pay for it. 24/7 they pay for it for 24 minutes, and then they pull it back down. Because they got to the airport, the Uber served its purpose. They pull it down to So basically they just, they just load the data up into the cloud, and then they pull it back down. They chew on it, and then they pull it back

13:21

down. The results out. You gotta Yep, and then there it is. You

13:25

got it. So they don't have to pay for massive amounts of compute and storage all the time in their own data centers. We handle that just like a crew. It doesn't have its own, let's say, hydroelectric dam to create electricity. They utilize the utilities. They're now usually utilizing the utilities, the industrial strength utilities of Microsoft's cloud compute, storage, and now AI that has the ability to do so many computational problem solving challenges. That's what they're that's what they're really doing when they're partnering with Microsoft. It's secure, it's hardened. It is it is very reliant, and it has the ability to do things we haven't been able to do before. Is there

14:04

a latency? Oh, always. But it's not. There's

14:08

always a latency. And it's this. It's how fast can you send light through through tubes? And so the closer you are even because you're moving at the speed of light, yeah, but even though you're moving at the speed of light, is the data compressed? So Netflix tried to solve this problem a long, long time ago, they created a company called Netflix, right? They were still mailing people's CDs to their home, to their to their mailboxes in their home, but they knew someday they were going to be able to compress an entire movie and send it through the internet. So Microsoft is working on, you know, working with compression technologies, all those kind of things, right? Yeah, well, Netflix is stored in the cloud. It's not on somebody's server, right? Absolutely. And then they pull it down. Yeah, there is latency, but that gets better and better, faster and faster as we're able to compress and decouple a lot of the stuff. We say, oh, we can't fit this piece through at the at the beginning of the movie, but put, at least put the first part through. And as you start. Watching it, the rest of its loading behind, but you never know, no, because they've organized it in such a way. Same thing with this. There's all there's always latency, but you have to look at it and say which pieces of information are the most critical at the beginning, you you surface those first, and then you let the rest of the problem come, come as you need it.

15:18

Brilliant. Yeah, yeah, that's brilliant. I like that. That's an interesting solution. How do you, how does Azure take into consideration everything's everything is growing exponentially. Yes, data, data, data, data, everywhere, getting set up. How do you, how do you Azure take into account that that's happening all the time? Well, I'll tell you that I don't know. I just,

15:41

16:32

it takes that happened overnight, yes. And we

16:35

cross big hardware and requires power optimization, so we're actually putting AI into power optimization for our very own data centers. Now show us how to optimize power better so we put it in the places where it needs to be, and we can also put our data centers in locations where they need to be in. AI is helping us make those decisions.

16:54

Yeah, it's dead sexy. Yeah, I don't know what else to say. You're the man.

16:58

Nah, it's our customers. You know, as they as the speaker said this morning, when you build the right kind of community, everything works.

17:07

How do they get a hold of

17:09

you? Go to LinkedIn. Jason, Michelle, M, E, S, H, E, L, L, on LinkedIn, reach out to me and we can talk. Fan

17:16

freaking tastic. Thanks, guys. A hell of a conversation. All right, listeners, we're gonna wrap it up on the other side. We're gonna have all the contact information. Contact information for Jason out on Industrial Talk, so reach out. You can tell he's got mad skills, most definitely, once again, a crew and insight is where we're broadcasting. It is in San Antonio, and it is a great community of individuals that are fascinating and wanting to solve problems. All right, stay tuned. We will be right back.

17:42

You're listening to the Industrial Talk Podcast Network.

17:52

Yeah, another great conversation. Jason, Jason Bucha Microsoft. You might have heard that company before we were at accruing insights, it was a great user conference with people like Jason. Jason wandering around. He's just wandering around. We're willing to dish out some insights to anybody who would listen. That's Jason. He's fantastic. Talking about how the power requirements associated with AI, it's all there. It's it's happening and it's gonna happen, and you better be a part of it to find people like Jason, sort of hang out with and pick his brain. All right, we're building a platform. As I always say, you have a podcast, you have technology. You want to be on a podcast, you just have to go out to Industrial Talk.com, Industrial Talk.com, reach out to me, and then we'll have a conversation anything to help you succeed and amplify your message. That's what we're all about. Be bold, be brave. There greatly. Hang out with Jason. Change the world. We're gonna have another great conversation coming from accruing insight shortly. So stay tuned. You.