 Good morning, everyone. Welcome back to theCUBE's coverage of Google Cloud Next 23. This is day three of a half day of coverage here for you, but you know that, because you've been following along on full days one and two. Lisa Martin here with Rob's Touching. We've got Pithian on the show next. Paul Lewis, it's Chief Technology Officer, joins us. Paul, great to have you back on theCUBE. You're an OG. Yes, I was been here many, many times though. One time, but it was in the original office back in Boston. That was good times. Enjoyed it. I know you were on the road with us. Talk to the audience a little bit about Pithian, the origin story, mission, vision. Give us that backstory. Sure. So Pithian 26 year data services organization. And Pithian is based in Pithia, which is the high priestess to the Oracle Apollo. Because we were a Oracle managed services organization to which we've now expanded databases and data and analytics and data warehouse and data lakes. I'm going to even expand into the application side and of course into the cloud because all data runs in the cloud. I think this is near and dear to my heart. I'm a data guy. I was at Amazon. I've been at a number of different storage and data companies. Like yourself, I did get similar backgrounds in that respect. Started out as a DBA very long, long, long time ago. But I think one of the things that really we're starting to see is not only CIOs, but the C-suite is starting to see the challenges. One of the big ones that we hear about and it's been, all the Googlers have come up and talked about it a little bit. But I want to see you expand on it is data security. It's key to everything. Trust and data security is at the root of this and how are you seeing that with your customers and how are you helping them with that? Well, we've seen a shift between, you know, infrastructure and applications being the IT asset to data being the IT asset. And now I have to think about not protecting the bad actors from coming into my network. It's about protecting the bad actors from taking my data. Because that's what I need to protect. That's what I need to monetize and therefore that's what I need to secure. And we used to think about it just being deep in the data center. It's so far in you can't possibly reach it. But the reality is data is now exposed right out to the consumer. So now you have to think about that entire chain of events on where to secure it. Where the consumer enters it, where we use it, where we trade with others. All of a sudden now each point in time I have to secure and then each part of the data I need to secure. Different people need different needs. Yeah. And how do you, how has the security landscape changed? The last few years we've seen tremendous change in the threat landscape. It's porous, it's amorphous. How has Pythian evolved to help organizations secure their data no matter where it lives these days? Cloud, Edge, data center. Yeah, excellent question. You're right, it is a very diversified deployment. It is, it used to be just traditional data center, then it went to SaaS, I used to public cloud and now I have private cloud implementations that I have third parties managing my own equipment and of course the Edge as you say. And since data is both originated consumed at all of those points, you have to secure each one of those points. So it's a zero trust implementation. Each transaction needs to be secured, each identity needs to be secured. And I have to ensure that the right people see the right data. That's almost more important than just stopping them from having it. It's that a data science needs to see the raw information but the end consumer only needs to see their information. That changes the filtering in many ways to support that. And then of course data protection, it's not just the primary source, it's the 17 copies. Yes. Right? It's the analytical copies, the backup copies of the archive copies. So that's a lot of sprawl. Yeah, I think that's one of the big things is data management and the whole how you treat data as a product is really where we're seeing it going and we've been talking with people like yourselves and others that seem to be pushing in that direction. But I think one of the things that is really key to all of this is the people and the processes and particularly in the people. There just can't be enough people to hire to go and do this. Everybody's getting pushed to do Gen AI, work backwards from the business value. What are you seeing in the recruiting, prompt engineering and recruiting and Gen AI? I mean, because now everybody's gonna come out of this. With all, I mean, you can't go through a segment without saying Gen AI about 15 times. That is true. In fact, in the preview, I go AI, AI, AI, it's gonna be everything. But how are you seeing the people part of that and the recruiting part of that going? Because that to me is interesting. It's fair to say that the skill sets are distinctly different now than there were even five years ago. So great things have happened. The pandemic created work from home. And even though there's a lot of returned office type talk, for the most part, it changed your recruiting from in the city to anywhere in the world. So now if there's a group of amazing data scientists in Lisbon, you can go hire them in Lisbon even though you're in Cincinnati. It's you don't have that barrier of limitation, therefore you get to see them. But prompt engineering is a great example or that's such a new skill set that it's hard to even find people out of academia, let alone people with one year's worth of experience. But to effectively implement Gen AI, you really have to have the psychology degree and sociologist degree to figure out how to best ask the questions in order to get the right responses. So now it's just not technology but it's technologies and humanities even more so than data scientists. That's fascinating. That is fascinating. But another thing in terms of skills gap that I want to talk about, let's go back to cybersecurity for a minute. There's a massive skills gap that has been going on for quite a while. A lot of organizations are trying, obviously, they're looking at tools like AI and Gen AI to help that. How does Pithian help organizations to fill in their needs, that cyber skills gap, can you? Good question. There is a skill set because there's more bad actors and they're better at what they do, right? The reality is the better they get, the more skills that I have to be able to defend that. And you're right, what Pithian does is we spend time first looking at your existing infrastructure to say, was it appropriately designed in the first place? So let's look at the foundation to say, did you really zone your data in a way to ensure that certain people only see certain data? And then we'll look at the schemas to say, are you segregating them enough that the business users and the data scientists see different data? And then we look at your end-to-end value chain, the zero trust implementation and we help you with the people and the process and the technology infrastructure to support that. And that's key, especially the people part. Exactly. Yeah, I was gonna say, I mean, we're at a technology company, we're at a technology conference, but still the technology to me is almost third on that list, on purpose because the technology tends to be the easier piece of it. And I think, you know, let's kind of shift a little bit because I think there's some other technologies that you guys get into more immersive technologies like AR and VR and things of that. What are you seeing? Because I haven't heard anything about that here this week, which is actually, now that I'm bringing it up, it's kind of weird as well because there was a lot of it, actually a little bit of it last week at VM Explorer, I almost screwed that one up, but again, but called it VMworld or whatever it is. But I think it's, as we go through this how are people really looking to bring that into what they do on a daily basis or how are they trying to monetize that or what is it that that technology's being used for? Good question. By the very nature of our business, we spend a lot of time in retail, CPG manufacturing, even healthcare on the hospital side where real consumers sit, but have very small IT as a general practice, right? So they have big, big brands, 10,000 stores, but 100 people in IT, right? So those are the types of organizations that require specialty skill sets for things like Gen AI and security and don't have the scale. They can't do three, four, five projects at a time. So in those sense, those organizations have much bigger digital transformation budget than they do an IT budget. So the CIO and those organizations are saying, I need to participate more in the business. I need to report to the CEO versus the CFO and therefore a lot of my projects are the immersive projects. So how do I look at the customer journey end to end in a retail? How can they start a transaction in the website but make their way to the store and eventually buy a product or multiple stores for multiple products? So that's where you see the augmented reality. The whole immersive technology is implementation where I might have a metaverse implementation for shopping, but that's just where I start or that might be where I get support or that might be where I read the manual, right? So I have to have all that world and then when I'm in the store, I have an augmented reality implementation where I can see the price tags and see the comparison studies with other products while I'm shopping. And you match that with loyalty. Loyalty is really the only way to increase the basket size in retail. So if you match the person as they walk in, the cameras are identifying, that's Paul Lewis, I can see what aisles are going down to. They use their phone to say, okay, the price on this is 323, but it's actually 310 in store B, which is five miles later, that's where they're gonna go. That customer journey end to end, that's the digital transformation IT money. Yeah, or you give them a coupon right there and then to keep them in that store. Yeah, 10 cents. I'll give you 10 cent coupon on that. And I think that's really the neat part of that. And I think you must be running into a lot of first party versus third party data and all the data privacy and security concern. I mean, it ties on to the security concerns is privacy and all of that. Are you, especially in retail? I mean, that's just huge with retail. That's actually the biggest concern with JNAI and fairness, right? JNAI by definition, as you've heard many times, is artificial intelligence to create content, text, images, video, but that's not really enterprise JNAI. Because enterprise JNAI has those concerns. What can I do with my customer data? Can I take my intellectual property, my algorithms and actually put it into a chat GP? Not probably not. Can I take code that's delivered in Bard and apply it into my core banking system? Probably not. So there's some real enterprise limitations. It's both complex and complicated. Complex in that I have a thousand points of light and complicated in that I have lots more technology to consider. What is the JNAI, given that context that you just described from an enterprise perspective, what does a JNAI strategy look like aligned with a data strategy? How is Pithee and helping organizations to really figure that out in a short amount of time? Good question. We always start with a data strategy that's near and dear to your heart. But we actually start with the data monetization strategy. And it's really six points. It's how can I package up data to sell it? If I can, clearly some companies can't. How do I learn about the customers or transactions or product more to sell them more? Or how do I wrap a non-data-centric product with a data-centric product? Like I can sell you a checking account, but for the low price of a dollar a month, I'll let you know when your spending habits change. We start there. We get a bunch of use cases and then we say let's prioritize those. Biggest bang for your buck. What's going to find customer segments you currently don't find now? Then we tackle the how. So what, what, what, what, what? Then the data strategy is where's that data? Is it inside? Is it outside? Do I need to augment your data? Is there an open source, LLM we can use to start to build those prompts? That's the process, right? And we want to do quick wins. Absolutely. You mentioned the great retail example in general. Is there a customer store you can share with us that started from that data strategy level and went through a complete AI strategy? Certainly. So we have a, we have a very big client called Cascades. They're well-known in Canada. They are a recycled fiber paper company. So they do everything from shredding and recycling, but they're most known for their consumer products, which is paper towels and toilet paper, right? And they're end-to-end. 80 plants. And what they have is all of their data stored in their ERP. And what they weren't able to do was create interesting insights for the consumers of the clients. So they had to A, figure out how to get data out of that, and then add sort of augmented technology to support creating value. So that's when we helped them create a enterprise data platform. And then we start to create machine learning implementation so that they can understand what a customer is and then start selling them differently. New baskets, new customer segmentation, even sentiment analysis with what consumers are thinking about that paper towel in general, right? Is it as good as others? That's crazy. That's very cool. I hadn't, yeah, I know. I hadn't thought about paper towels and sentiment analysis of paper towels. Yeah. But it makes so much sense though. It does. I have strong feelings on paper towels. That's actually how they figured out to build a product line that was eco-friendly, because that's what the consumers were looking for. You know, you've seen paper towels where you get to take just half of them off? Yeah. That innovation came from. Those are my favorite. Yeah, exactly, because you're using less and therefore, you know, better consumer. Yeah, I think that to me is one of it. It's working backwards from that data product. What is the data product in the data app that's going to be used or consumed by the customer, which could be internal or external from that. And I guess you guys see that a lot throughout all of this about helping them, again, with the strategy first and moving through that. Is that typically how one of the engagements goes? You're really starting at the strategy level and helping them. Like you said, let's see what you have and understand your current strategy and where you're trying to get to and the outcome you're trying to get. It's actually very rare they would have a data strategy. It's much more likely that they've invested, you know, 10 years and $10 million into a data warehouse, because they wanted to build reports. They've had a thousand amazing reports. And we'd never go back to them and say, you know what, that was a poor design. Because it's perfectly good for the 10,000 reports that they have. What it's not good for is all the other data that they have or they want externally, all that unstructured information, all the streaming data from the cars. It's limited to terabytes instead of petabytes of information. And it has to ensure that Paul equals Paul equals Paul, but that's not how the external data looks. It looks like BW Lauer. It looks like Darcy Lewis, which is Paul Lewis's husband. It is a mishmash of complexity to which I have to deal with new, interesting cloud technology. And I would put everything on the cloud because that's where the innovation is. It's Google that creates machine learning innovation, not software built inside your company. So, GenAI is cool and sexy. We hear about it at every event. Like you said, you can't have a conversation these days without mentioning it 10 to 15 times. But organizations have to be ready to benefit from it. On the people front, we talked about people process technologies. How do you help the people from a cultural perspective be ready for it? Because if they're not dipping their toe in the water yet, as everyone's saying, they're already behind. Okay. Fortunately, the last couple, three years, the role of the CIO has extended dramatically. They used to just be keep the lights on, red, green, yellow projects, blinking lights in the data center. Now they're expected to deliver the digital transformation portfolio. So for the most part, they're looking at their team saying, how are you going to help us find 20% growth in our customers? How are you going to help us introduce a new physical product, a new design for the car? And it's now up to IT. So they're looking at the skill sets of their teams to say, how do I add skill sets I don't have, like prompt engineering or data science? Much more difficult, we're in the middle of the US, right? We're there in Wawa, Kansas. They're not going to find a $350,000 data scientist. They need to really look at what we refer to as ecosystems of partnerships, of talent, of technology, of information. And if they want to scale or add skill sets they don't have, they're going to work with partners like Pythian to find that. And if they want to do five projects instead of one project at the same time, they need to scale. They need team two and team three and team four to make that happen. Now that makes so much sense. In like the limited time we have left here last minute or so, what are your thoughts, you're going to walk away from Google Next 23 with? I think Google Next 23 is better articulated instead of Google Cloud Platform. It's Google AI platform. While you didn't hear a lot of cloud in the keynote and others, it's because that's now the assumption. It's now not a migrate to cloud, it's now use what's on cloud. That's the interesting part. And the use what's on cloud is the AI, right? It's not necessarily gen AI, that's some of it, but it's just machine learning, right? It's just use your data for the purpose of something other than what it's original purpose for. And that might be just simple monetization. Totally agree with that. I think that's been what we've been seeing is it's about the whole solution, not about the IaaS or the PaaS or this service or that service or 1800 services or something like that. It's not about the individual GPU versus CPU. Yes, although we have the TPUs right behind us. We've been staring at them and we geek out on that stuff too, but that's a whole different story. Well, optionality is good, but you also want packages. Absolutely. Solutions, I think that has been the thing we have come away with. They've really done a good job of pulling the thread on the solution throughout their entire stack. And I think that it's been a home run for them. 100%. Speaking of home runs, what's next for Piffian? What are some of the home runs that you're looking to hit? So our focus now is both workshops with our clients to really appreciate their monetization goals. So we create a series of seminars to allow them to get a better understanding of what technology and strategy offerings are available. We'll walk them through their use cases, their monetization goals. And then we create a roadmap for them. We want to do that on security. We want to do that on GenAI. We want to do that just on simple data analytics. Because for the most part, especially those retail CPG manufacturing, they're very, very early in that stage. And they'd love to get to a point where they're starting to just see insights. And we can start there too. Awesome. Paul, thank you so much for joining Rob and me on the program. Great to have you back after a few years but telling us about Piffian, what you're doing, how you're helping organizations really be strategic from a data perspective. We really appreciate your time and your insights. Thank you. Yeah, thanks. All right. For our guests and for Rob Stratte, I'm Lisa Martin. You're watching theCUBE's day three coverage of Google Cloud Next 23. Stick around. We'll be right back with our next guest.