 Hello, welcome to the special CUBE presentation here in Las Vegas at Google Next 2024. I'm John Furrier, host of theCUBE, accelerating innovation with persistence. I've got two great guys. Oyaan Banerjee, Chief Strategy Officer and Growth Officer at Persistent and Ray Wang Founder, Chairman and Principal Analyst of Constellation Research. Ray, great to see you. Oyaan, good to see you again. Great to see you as well. So enterprises are continuing to evolve with Genervai, that all the hot action on the chip side, certainly the hyper-scales on the consumer side, all the social apps, large scale. Enterprise moving fast into AI Ray. The data we heard, Thomas Curzon, ground your data with your enterprise data. This is a big part of the enterprises. They can't stand still, they have to move now. Yeah, that's a great point. Every company is running out of data and the question is really, do you have enough data to achieve a level of precision that your stakeholders are going to trust? And we often use this example at 85% accuracy in customer experience. We're super excited. We're like, great, this is going to be amazing. 85% accuracy in supply chain means you don't have a manufacturing operation. 85% accuracy in, I mean, finance, someone goes to jail. 85% accuracy in healthcare, it's not enough. And so it's really important to make sure you have small language models, very small language models to complete that level of precision and data. But start with the data you have and then partner for all the other sources. Your team at Constellation is doing a ton of research. We've been tracking a lot of the stuff you've been going after. A lot's coming out. What are you finding, because some of the data we're seeing here at Google Next is super excitement. The confidence level is getting there, not 100%, but there's still a lot of prerequisites needed to fully have general AI. What are you seeing in the research? What's your research telling you? Where are the enterprise at right now? Could you scope that up for us? Yeah, we completed a number of AI surveys and I think the most important part is this year everybody got budget. Last year they stole budget and tried to make something happen. This year these projects are funded. And what's important is to start with your data strategy. You've got to have the data down, understand what data you have. And then what you're really focusing in on is having the right governance principles and then also making sure that you know what data is missing so you can answer the business questions. You start with business value first, understand those business questions and you can figure out the rest on the fly. Oh yeah, I want to get to it with the customer side. You guys are engaging with a lot of the top customers. What's their mindset like right now? Obviously they're experimenting, they got budget as Ray pointed out. So they have a mandate. Go figure out the bridge to the future. It's generative AI. Clearly there's no debate on that. But what's the first three steps do you take? What are they thinking? How are they applying and crafting a strategy? What are those first three steps to get there? Yeah, look, I mean, we're behind the clock six months and it was all about saying, what's your generative AI strategy? 150 use cases. Talk fast forward to today. It's all about saying that 150 use cases is absolutely no consequence. But look at where is it that you're going to be deriving the maximum business value. And it's really coming down to the two, three, four of them. And the question is, how do you take those two, three, four established feasibility, drive certainty? Your 85% example is a classic one. And the question is, where do you really have the maximum bang for the buck? Also picking up on your common Ray from a budget standpoint. I think so what you're going to continue to see is that budgets are going to be tough. I mean, as much as there is money waiting to be deployed, but that is going to be very selective. So again, question is, companies need to be very deliberate and the conversations we are having right now is moving from saying that, why don't you help us establish BOCs across a whole bunch of things to saying, here are five where we've already established a degree of certainty. Help us take it to production. Before we get into the, I have originally IT services market. I want to just quickly get an example. A lot of naysayers out there or people who are afraid of AI. Oh my God, they're going to, Skynet's going to take over the world. Oh my God, we need regulations. AI's got to run a little bit. We all, I'm pro-AI obviously. I want to see it. We got to watch the guard rails. But most people are like, I don't really see it. What is Genevieve AI? Can you guys give a proof point that would be so obvious that say, okay, that was not possible pre-Genevieve AI. What is Genevieve AI? What does that actually mean? And give an example of an obvious thing that would convert someone a naysayer. You want to go first? Oh, you go first. Definitely. Okay, yeah, look, I mean, think of the life of a software developer, right? A software developer or a software development team was coming up with a new release on your Apple phone every two months, right? I think that is now going to get shrunk massively. You're going to start seeing that Genevieve AI is going to come in and make, you know, the life of the software developer much more productive. You're going to have agents coming in and assisting them in their entire software development lifecycle. And you're going to start seeing the shrinking of that. You're going to start seeing more, you know, error-proof software releases. It is happening. It is no longer a figment of imagination. So to all the naysayers out there, we can look at the 85% example that Ray gave and say that, oh my God, 15% is still bad. The fact of the matter is this is temporal. This is today. You know, I mean, the same race three months later, this 85% will look like 98% and counting. Ray, what example would you give? Because, I mean, you see a lot. But to someone who said, okay, give me a business example of that's so obvious and I have to believe in 100% and go, I had 120 use cases. That's good. We're down to one, right? I think the one, number one use case we see is procurement, right? And why is procurement so important? You throw all your contracts in there and you're like, oh, wait, we're paying this guy this rate. We're paying this rate over here. We're not getting our terms of conditions that actually match over here. And by the way, oh, we missed all these regular requirements that we were supposed to do. All that surfaces up. That was like thousands of pages of contracts. You had to read. Suddenly you can just ask the question, hey, which one of my suppliers is actually getting paid earlier? Which one of my suppliers is actually in compliance? Do we actually follow our export rules and regulatory requirements? That's why it makes it so powerful. So what would have taken you a year and a project? You could do that now in a day or less. You know, interesting, Thomas Kurian's keynote here on Google Next, he talked about the Vertex AI having 130 models. One of the things he mentioned is the context window of a million tokens. You can put 700,000 words. Okay, that's a lot of contracts, Ray. So you're envisioning, they're just ingested in and all that is reasoned on behalf of the user. I think you can ingest patterns of contracts. You'll say something like, oh, see like those patterns actually emerge. And then what you'll be able to learn from that, is hold that back into the model, train, and then you'll want to use them as tokens on the second time. You're going to get really good at skipping words in those contracts because, you know, 90% of those words are useless. So you'll actually optimize that training down to the point where you only need like 10% of the contract and you'll be super token efficient. Let's talk about IT services, the business that you guys are in. You guys are doing well. A lot of the folks in your market are helping customers cross that, build the bridge to the future. Never mind crossing the bridge to the future. And so I think it's important to note that this is an experimental phase formulation stage where there's real advisory, real work. Okay, so I have to ask, like where is that going? And because at some point the IT market has to evolve too. So okay, you can teach them how to fish, so to speak, but at the end of the day, IT has to get to the table in a new way which everybody, what's going to change in IT services? If things like procurement happen, like Ray just pointing out, I mean, this is going to be operational efficiencies in IT. And are they ready even to step up to the plate and lead and execute? What's your take on the IT services market? I'd love to get your perspective as a researcher. I think the one fundamental shift that we're seeing is we are no longer talking only to IT. We're starting to talk to IT. I mean, pick a large pharma company, right? So we're talking to the chief medical officer. We're talking to the chief product officer. We're talking to the chief data officer. We're talking to the chief marketing officer and so on and so forth. Why? Because the problem is no longer about IT. The problem is about taking a business problem of getting a drug out to the market in the shortest time possible. It's about shrinking that stage four of your clinical trials. It's about finding the right set of patients to be able to actually have them stick and then be able to establish getting the drug out in the market. So I think that the foundational shift that we are seeing right now is, as much as the CIO organization or the CDO organization is still the anchor for players like us, but we are starting to really massively get multimodal with our clients. And it's all becoming a question of what business problem are you solving and what is the technology stack that you're bringing to bear? And in today's day and age, it's all a partnership model, right? So I mean, today the work that we do with Google and it's been going on for a decade right now, it's really saying that how do you really get together with an ISV, with Google, with us and get a complete end-to-end solution towards that line? That's a foundational shift that's happening in the market right now. Ray, as an analyst and your team of analysts are looking at this now, we have disruptive technologies hitting the market at big time. That's going to put a change in the IT services market. How do you see the services changing and how do companies then pick up on their side? What's your analysis? Yeah, I mean, I wouldn't want to be a BPO player right now. I mean, BPO shouldn't exist, right? That's a market that should be cut by one-tenth, right? But it's not really just the automation and the AI that's going on there. There's other work that can be done and the optimization actually happens once you digitize a lot of that work. And so we've gone from figuring out when do we have full intelligent machine automation to when do we actually augment machines with humans? When do we augment humans with machines and when do we actually put humans into work? The design point's always been automation. The flip that's actually happening right now is actually going to be when and where you insert a human in the process. And that's going to change the way we actually design contracts and design services. And so if you're going to build BPO in the future, done, we can actually software ties BPO and actually believe like all those BPO players are in trouble. When do you know that you're facing irrelevance? What are some of the signs if I'm a company? You know, the frog in boiling water, as they say. And because, you know, right now there's two sides of the street. There's the right side of history of NGI and the ones that aren't a GNAI and the ones that aren't going to move fast enough. So when do you know like, I better either pull a ripcord here or I better make a change? When do clients, Ray, face that moment? Because we're kind of starting to see the beginning phases of people who if they're not acting and putting things in place, they're going to be quickly on the outside looking in. Well, there's only about a hundred companies that are really doing this right. And what I mean by that, but take your example for molecule to market, less than a billion dollars, trying to compress that to about 500 million. We're actually facing more than just this AI arbitrage that's going on. We're actually having a margin compression. And this margin compression is something we haven't seen before. I'll give you an example. Like, do I need to go buy all the CRM software out in the world today or can I take a player that can deliver at one-tenth the cost? You're going to see one-tenth the cost in almost every single interaction. So we have margin compression happening in the world's biggest deflation about to happen with some of these technologies because of AI. But we also have demand on the other hand because we're going to do more. And so that's the other side that people forget. So we're going to see that kind of model actually emerge over time. Oh yeah, this is talking about wealth shift. Talk about the value shifting other places. This is an opportunity for companies to move fast and get on which, how do you see this in the policy growth, you're in growth in chief strategy, how do you see it in growth? This is in your wheelhouse. Look, I mean, this is going to be our calling card. I mean, let's just put it straight, right? I mean, persistent. I mean, I think the main persistent comes from persistent data, right? And we've kind of lived this now for the last 33, 34 years. And I think so today, if you think about the next five, it's going to be all, whether it's gendered AI, whether it's vertex, you know, you talk about whatever that's going on from an overall business transformation standpoint, it's all going to be housed in the data. And I think that's where we see it to be a phenomenal opportunity, where we go and help our clients be able to really unlock the full potential of gendered AI. Let me give you one more thing, right? We had artificial refrigeration in the 1920s. You're like, oh, is it going to be hot, Kelvinator, Frigeraire, G, Plastics, Plastics, a different generation. This is in the 20s, right? We weren't there, nobody was here. But the point being is like, we, this was going to be the revenue share market, like cold stuff, right? Frigeration was going to be hot, rooms were going to be air conditioned. It was going to change the way we work. Who won? There wasn't any of those companies. It was Coca-Cola, because they got cold food supply chain down better than everyone else. And that was the innovation. Out of those hundred companies that are in there, one of them is going to make that breakthrough, like that, and then transform the industry. And that's actually, then everyone's going to follow. And that's what we're waiting for in this thing. This brings up the opportunities ahead. You guys are seeing a lot of customers that are looking at this picture saying, where do I fit into this? The future's unfolding right in front of our eyes. There's a lot of opportunities. What advice are you giving your customers on how to seize them? I guess, first things first, we tell them that, you know, be very deliberate, you know, focus, prioritize. You know, don't try to do everything all the time, because A, capital is not infinite. B is you will have to establish your creds, rather quickly. You take a few wrong steps, and that is it. You know, you're not going to be advancing any further. So we would much rather go tell our clients to look at, you know, a handful of things and take it to production and establish the value before you start really taking it to the next level. I mean, it is counterintuitive. It is counterintuitive because I would want to do much more and sooner, but the risk always is that if I try to do much more and sooner and then have a few missteps, it creates a big reputational issue. So it's really about being deliberate and thoughtful about what we are actually doing. And taking a portfolio strategy, right? Return on transformation investments, really important, right? And what you're thinking about is, where are my regulatory projects? Where are my cost reduction? Where is operational efficiency? Where is revenue and growth? Having the right portfolio is going to be very important. Ray, when you talk to customers and they ask you about IT services, the role persistent plays, and they look for your advice on how to shop, I think about my engagements in this new era. Because you mentioned this, you can bring margin compression down here. You can buy by the drink, buy by small packages. Remember, generally, you can re-stitch together new systems who are starting to see people do that. I mean, a little bit more advanced, but I can envision a future that's going to happen. What do you advise them on how to handle their engagement with IT service providers and people who are helping them build the next generation? What kind of advice do you- I think it's very, very different. They're IT service providers that have a full breadth of services around the way. And then there's persistent, which is focused on innovation, product development, engineering, right? They're building things for companies like that. And so I don't see them as the IT services that you're just hiring people for, for, you know, timing materials. I don't see them as someone doing that. I see them as people you go to them for an innovation partner. And that's kind of how we advise our customers. Well, bring that up. I keyed that up. I want to double down on that because we heard on theCUBE here at Google Next, someone came on and said, in the data equations flipping the script because you need data to run AI. And the old digital transformation was give the data to the BI team or the data scientist. Now let's give it to the computer science team and the software engineers. It's a system discussion. So to your point, you guys are doing a lot on the development side. This is, I won't say a new formula. You guys have been doing it for a long time, but this seems to be a critical element. I love your comments too. As clients have to think about their business as a system, almost a system software layer with a bunch of gear now, new gear. It changes the game. It's the same game, but now you have new stuff. Yeah, I mean, look, the way we are seeing it is that everybody the whole world was talking about generated AI for the last 15 months. But we believe that generated AI is just a tool in the toolkit. So if you really think about the stack, the system is bigger than AI. AI is bigger than generated AI. You have to look at it from a system standpoint to be able to really leverage the full potential of this entire movement that's happening right now. And your comments, what's your comment on this? What's your reaction to my statement that it shifts to software, not just the BI team and the data science team? Oh, it's all software, right? We're just on different stacks. We're actually applying different business models to software. Software is Lego blocks today. We can just mix and match. The only thing holding us back is really our ambition. We're going to be able to build these things super fast. We just don't know which ones first. And I think that's what we're going to learn. What was once table stakes, the next table stakes level comes up. When does that human creativity become table stakes? Yeah, but before that, I mean the comment that we made, right? I mean, today an automotive company is calling themselves a software company, right? And the reason is quite straightforward. The electromechanical stuff isn't changing much. But how do you get the maximum juice from the electromechanical stuff is all software. So I think that that's really where someone like us come in, because to raise comment, you're not the traditional soup to nuts IT provider. But what we do is really play more in the CapExide of things. We help companies build products. And then we take it to the OpExide, which is actually run and sustain them better. And to that extent, we are building a whole bunch of accelerators, connectors, standardized platforms, which we take to our clients, which helps them accelerate their overall ability to infuse JNAI in everything that they are doing. You guys really come in and help them figure out the future, get them up and running, pass them the keys, let them take it from there. Give an example, if you don't mind, because I think people are kind of coming to this realization that I need to have this systems approach. I got to start building. What, give an example of how you guys solve that problem. A lot of enterprise wanting to play with a lot of, you know, what I would call as POCs right now. And every time you do a POC, you have to do it in a very secure environment. You have to look at your data and publicly available data on LLMs. So what we've done is we've actually built a very common platform. We call that as the GenOS platform. And think of it as a completely prefab integrated, you know, suite. The whole stack is integrated, where we have our clients come in and start really building up those LLMs, start training those models, and start establishing feasibility on top of that. It is not locked to any hyperscaler. It is not locked to any LLM, but you can actually do whatever you want to mix and match and be able to get the most out of the system. So there are many such things that we are doing on the back of, you know, enabling our customers to accelerate their overall journey. Well, it's a great business opportunity for you guys right now, a great climate for your business. Ray, we're here at Google Next, and so you see what's happening, you see the infrastructure, got to get more chips, got to get the custom silicon, got to get the arm thing going on, relationship, the processors, you got to get the horsepower. Then you got the user experience side with Workspace and everything in between. Those are the two key areas because everyone wants that generative experience. What's your take on Google Next this year, if you had to kind of boil it down to, you know, where they're at and what it means for customers as they look at this future? I'm actually very impressed. This is one of the best shows I've been to all year. Nvidia is probably the number one show, this one number two right now. And part of the reason it's not just the announcements or what's going on, it's the death of the ecosystem. We walked that show floor, it's everybody there. We spent like some time with the president of operational technology at UPS. He's out there, right? He's talking about how they're taking RFID tax, actually digital twins and putting it together. So the conversations are here, the excitement is there. And then of course, we're seeing a whole bunch of activity just from partners and also individual companies that are actually talking about what they're doing. And that gives Google a good opportunity given the markets changing right now. Whenever you have the ground shifting, people can lose their position of the incumbents you mentioned, the refrigerator example, like none of those guys won. There could be a new entrant coming into this market, a new brand that no one's ever heard of that could give Google a big lift. Yeah, you and I know those names so we can't talk about them right now, but actually it's very true. Thanks so much, closing statements. What should people know about this year? Google next going forward is that the year of agents, the year of internal twins, external apps, how would you summarize this year going into the rest of the year? Year of developer productivity, year about accelerating product engineering, and year about making it real for the organizations. Yeah, well said. I would say, look, this is the age of AI. You're going to start today and you have no idea what's going to be in the future, but if you don't start now, you're going to be behind. I'm glad you're behind. Great, thanks for coming out. I really appreciate you all. Good to have you to see you out again. All right, this is the CUBE Special Presentation Accelerating Innovation with Persistent. I'm John Furrier with CUBE. Thanks for watching.