 Good evening, everyone. Welcome back to theCUBE, the leader in live tech coverage. We've spent the last two days unpacking everything from UI path forward six here, MGM Grant in Las Vegas, Lisa Martin with Dave Vellante, Andy Chariah joins us, the VP of Consolation Research and Rebecca Knight. Guys, AI at work was the theme of the event. I'm going to start with you, Andy, as our guest. What were some of the things that you heard and saw and that impressed you about where UI path is going? Okay, for impressiveness wise, look, until now, they were more about more of a UI automation and some rudimentary process automation, but now they're trying to move to up the value chain, trying to automate some of the AI things we talked about it. They were even throwing the term around calling AI automation, which is a uncharted territory, if they can figure it out truly. I think that's a huge opening for them. Do you consider UI path an AI company? You follow AI companies very closely. You don't have to ask about Anthropic and Cohere and Huggingface, they're AI companies, you'd agree. Is UI path an AI company? That's a tough question, man. Why do you have to put me in this situation? Come on. All right, so. Well look, this is one of the top AI analysts in the business. Oh, thank you. I appreciate that. I appreciate your comments, thoughts on it. So is UI path a true AI company? So it depends on your definition. It's more of an automation company with some AI capabilities, is how I would answer it. Not a pure AI company, because they don't have a lot of abilities that I would look for in a pure AI company. Like what? Like what's? Well, so to begin with, a true AI company should have mastered their data. So got to have a ton of data to train your AI models and you should have the ability to train your own AI models and you should be able to make cognitive decisions based on that. So there are a lot of elements in there. That's interesting. So what is UI path sort of data strategy? It's a little bit, I have to say, maybe because I haven't done my homework on this, but what is their data strategy? Your point of view. I mean their data strategy is good enough, but the problem is again, keep in mind they are not a data-based and model-creation company trying to do things on that. So it's more of an automation thing. So they have enough data and automation, process automation, UI automation. But what should they be there for? I guess Teradata here. I was just surprised that Teradata's in the ecosystem here. I haven't had time to stop by. I'd like to learn more about why they're here, but it's not like they have an analytics database. I don't see partnerships necessarily with Snowflake or Databricks or they could use Postgres or Mongo. I don't know, the data is where it is, right? And it flows through the system, but it's not really captured. Yeah, but it's, again, they are not in that business. That's my point, right? I mean, in order to do that, in order to create, again, then your competition changes. In order to truly become an AI company, then you would do those things, but then if you do that, then you're playing field changes. I mean, the reason I'm firing these questions is I'm trying to think, okay, what's next? I mean, when I have talked about this, this company was, I mean, they were a rocket ship, 38 billion dollar valuation and then the IPO and now it's, you know, everybody's down, they're down, but they have such potential. So my brain is going to, okay, if we had to advise them, which we do sometimes for companies, what would you do? What could be the big move to take them to a 10X valuation? And I think a data play. No, that's a competitive field and I would, no, so I did have a conversation with their CEO and the chief AI officer today and some of the things came up in the conversation, but at the end of the day, going back to your valuation, the reason why they were valued at that time at that high is because of their potential when they came out. Ultimately, everybody thought that they were going to kill it, right? And then Microsoft landed up buying Power Automate and you know, there were other competitive players who came in the market, but then also the market is somewhat saturated with all these players going for this many years. They have figured out and finished almost all the low-hanging fruits and that's why I was saying earlier, they need to move up the chain to provide some value in finishing the AI-based processes, which seem to be very manual nowadays. So if you're able to insert yourself in the value chain, again, going out to the data platform, creating your own data models, doing things on that, would be time-consuming, but more importantly, it could be expensive as well because the AI engineers are cheap tire. We discussed it over, you know, on the other day. So do you want to really get into that market? I don't think they can chew that bite, so probably not. I would say, and I'm dominating here, you got to jump in, you guys. Sorry. I would say this. I would say I definitely would not want to see them go compete with a snowflake, for example. I think it'd be a disastrous mistake, but I do think they could become a new type of data platform where it's like, I use the Uber for the enterprise as a metaphor. People place some things in real time, data coming in, and systems of agency with AI taking action. They're not in a position to do that today, that would require invention, or partnership. I don't see that as a value, but there are a few other things we talked about in today's strategy session with them. So a couple of recommendations I made. So again, in my view, they had a play into this trend. One of the strongest points what they have is the document understanding. Not many companies, I mean, I'm not talking about OCR type of scanning the documents. I'm talking about really understanding the documents. The intelligent document processing now goes to the next level of document understanding, which means you're able to converse with the documents, which means, for example, if I want to do something, for example, my travel policy, remember they were showing an example of the expense report policy instead? So if I'm traveling here, if I'm not a big spender like dropping 500 bucks for a dinner, I want to know what my company's expense limits are. So if I'm able to go and ask a question that it'll respond to if I have an integration with those systems, making the process smoother for the employees, for the processes, and whether it's expense or any other way, right? And that's the demonstration they made. My view that the naming is somewhat misfit in there. It's not an iPass, what they have. It's more of an integration using APIs or API process automation. That's what they have, which is what we talked about. So they could potentially win that race, because they come from a different angle to integrate with all those massive systems which are exposed using APIs, versus going and fight against companies like Boomi on the other iPads, that's our even data companies. That's going to be a big fight. I don't think I can end that. I have some thoughts on that, but I'm going to bite my tongue. You've got to do that. What I'm really curious to hear from your perspective is where are we on the time spectrum of how AI and automation are really transforming the workplace. This is, as Lisa said, this is the theme of the conference, AI at work, and we've heard from lots of different companies how they're using AI and automation to make their processes better, more smooth, more efficient, make their employees more productive. We also know that AI's been around for some time now. Of course, it's not new, but yet the advent of the chat GBT and other LLMs makes it feel pretty new to the normal people, at least. Where are we in the development, would you say, in terms of how AI and automation are really going to change the way work is done and change the way the enterprise operates? Right. I would, in my view, I think we're barely scratching the surface. Because if you think of your... So we're newborn stage, baby? Not even there, right? So, first trimester. So, going back to your point about AI's been around for a while, the AI term was coined in the 70s, and there were a lot of things that was actually... One of the reasons why AI's coming out at us so fast is because all of the, what do you call it, the ideas, the theories, all of them were well-baked for many decades. 30, 40, 50 years, right? It's been already ready, but there was a technology limitation. You couldn't scale up your compute power using cloud. You don't have enough data storage. You don't have the GPUs, and you didn't have the AI engineers to think that through, and you didn't have the infinity band, big network of data movement. So, when you combine all of them, there were so many limitations that you couldn't think about the massive scales of the chat GPT level foundational model. So, they were doing, as they call it, as a Mickey Mouse small model. Hey, what's your hour? Yeah. Trying to make it work with that small model, so it had so many limitations, so it couldn't work. But now, the sky is the limit. I mean, it's truly, when you use cloud and the GPU and the combination to expand those things, it's literally, you can build massive size models. I mean, the chat GPT-5 they're talking about, I couldn't even read the numbers and how big it is. We are talking about a huge number, so that's all I could say. I couldn't even remember how much it is, that much of a parameters and tuning that it could do. And you basically boil the entire ocean of all the data that's available to the human kind until now, and you're able to figure it out using that. But having said that, I think one of the things Dave, you and I talked about is, would AI ever be ready to handle the unknown unknowns? Yeah. That is going to be very debatable, because when it comes to, in my view, when it comes to human mind, where we excel, one, of course, is laziness. We want machines to do everything, right? Two, if we encounter something that we haven't seen before, a human brain, most ways, would be able to figure out what to do with that situation. That's where AI is still not ready yet. It's stuck, saying that, you know what? I don't know what to do with that. So that unknown element, it may not be able to handle. But for the known things, what it can do, the art of the possible is almost unstoppable. So is it that it lacks curiosity, or what is the attribute that we need to give AI in order to help it figure out the unknown? So as one of the CIOs that I was advising put it succinctly, I've been using this technology for many, many years, you said, and I've never been so excited and so scared about the same technology at the same time. Yeah, which is... I'm a CIO's mouth, that's unnerving. It's like the art of the possible, because you're looking at doing things like in the most cost-efficient way, the most easiest way, the art of the possible of things what you thought is not possible, integrating with certain systems, doing a workflow in a certain way, or producing content in a certain way, or even producing code in a certain way. That is all limited by our limitation and thinking powers and whatnot. Now the machines are assisting the humans to do that. So the capability is unlimited. But then with great power, as they say, comes the responsibility too. Most of the AI today, they're not learning systems. This is perfectly why I think full self-driving has been such a challenge. It's mostly knowledge systems. It's not expert systems, it's almost knowledge systems of it knows the information, you already told them this is what it looks like. So we're trying to put these things together. Why is it that we don't get to drive until we're 16 years old? Because the cognitive learning occurs. And so then the question is, okay, when will we have these learning systems? And it is scary. I mean, you know. Well, to your point about, the G.I. is like, uh-oh. Oh, I have a whole another idea on that. But talking about a fully autonomous driving systems, I posted on LinkedIn, I don't know if you saw that or not, we talked about it. So Tesla, there was a Tesla that is going in and then it encountered a horse drawn chariot in front of it, couldn't figure out what it is. It was trying to flip between a truck and a human being bicycling or walking and it was automating the speed based on that. Imagine if you had to speed up based on something or speed down or something, if you had to make an erroneous decision which cost lives, that could be dangerous. I mean, in that case, it worked out okay, but that's the problem. Well, and you think about how Tesla works, right? I mean, it's looking at lines and it's looking at red lights and green lights and it's essentially looking for these patterns and then saying, okay, go, three trained images that it has seen before. To your point, we've got to get to the point, what happens when you go off road or there's no lines in the road or it can't handle well. So it can't, but it's collecting a lot of data, it always comes back to the data, doesn't it? Where in all these videos and images, which I would think gives Tesla an advantage relative to some of the other, you know, automated vehicles. Look, AI is, you saw the chart that they put up on that this one we talked about it on the Ed Chalice talk, right? I said, it was an Ed Chalice talk and he had, I just wrote it down. It really compressed language understanding, math, code generation all came within like three or four years, like out of nowhere. Again, it's because a couple of reasons. One, we didn't have enough data available until now to do those things. But like I said earlier, it had the capabilities and the thinking behind it's matured for 30, 40, 50 years, right? But he couldn't execute it because there were technological limitations. Now, you got all the theory in, you got all the technology capabilities in and that's why the innovation is becoming so faster. And they also said GPUs, but what they didn't say is somebody had to figure out how to program those GPUs to take advantage of all the parallel capabilities that haven't existed in traditional architectures and most people can't get these super powerful GPUs. I mean, they're either too expensive or not available. So what's going to happen when two things happen? The cost of these things comes down and they become more widely available. I mean, a whole new set of applications is going to be developed. Well, so yes, if you're purely dependent on GPU, right? But there are other players in the market. For example, Sambinoa's developing chips that they want to compete with and there are three or four startups that have briefed me based on just some of the capabilities alone. And Arm, come again? Oh yeah, and Arm, no, we're going way longer than that. He's saying two more minutes. Come on, we got this AI expert here. Oh, oh, no. Okay. We're going to pick his brain all night. Oh my God. Pick his brain and roast it. You're too happy, Andy, you're happy. Yeah. No, so that's the point, right? GPUs used to be the only game in town, but it's not. I mean, Google is developing their own chip and Amazon is trying to do the same thing. Azure is rumored to be doing that too. Everybody wants it, you know, playing on it. Arm is going to do it. But the most immediate one is, you know, Sambinoa systems has a pretty decent system that I've seen and then there are a few other startups that are well funded right now and they showed me some capabilities while more numbers, metrics, on what they can do compared to GPUs. I mean, there is a, everybody wants Nvidia to have some competition, no question. Well, so at the end of the day, look, Jensen Wong is one of the best storytellers ever. So it's more than the GPU, it's about the story, what do you say else? I mean, was GPU and Nvidia a bonus AI company? No, it was, what were they doing? Graphics, acceleration, cards, video games, and then crypto, and then he moves the story along. Something else he'll find. But they do have a lead in terms of software, architecture, it's well embedded, baked into more ecosystems, but there's a mighty worthy competition that's coming up, is my point. Speaking of stories, I'm curious what you heard from the voices of customers the last couple of days, Andy. One of the things that we've had the great opportunity at Rebecca, Dave and I on theCUBE to talk with a lot of their customers. Yesterday's keynote started off in a very dramatic fashion with customers talking about the impact that they're making by partnering with UiPath. What were some of the conversations that you had with customers based on what was announced, where they are on the automation journey, where they are with AI, and a total fee, but that really hurt. Well, first of all, if you spoke to that many customers, you should be telling me. Not the other way around, right? But one of the, so I met a few customers and every single one of them, the way they were able to optimize their, not only processes, but more of a cost optimization that they're able to do using those things. And the amount of savings that they were able to do, and the way they are able to savings pass on to their customers, that's unbelievable, right? I was also talking to an S.I. early on a few weeks ago at their conference and they made the same point to the code generation, the software engineers doing code generation, one of the most complex things even now. So for AI LLM to figure out and do code generation, there are, I'm talking about 18 to 20 tools on that front, but of course GitHub is the leader on that. For them to improve the efficiency, S.I. is 50%, that's going to save a ton of time for all these clients, right? It's not just for UiPath, but overall I'm talking about. So you got to figure out a way to have a working model you will plan on passing those to the customers, you know, at the end users. Definitely. I want to ask you what you make of the citizen developer movement, because in some ways it feels inevitable, but I think about it from my own perspective as a pretty non-technical person, the idea that I'd be writing code is pretty mind blowing, but yet when you think about Gen Z, who will be fully enmeshed in the workforce in just a few years, yeah, I buy it. I buy that they'll all be doing coding as part of their jobs. Where do you think we're going with that? And are there any concerns that you might have about a lack of security, governance? So that's not new, right? I mean, when it came to mobile development, the same thing happened, you know, everybody became a mobile developer and whatnot. So one of the things with the citizen developer thing, particularly with AI, with a lot of those tools that's being provided, it's a very dangerous, sharp, double edged sword that they're playing with, playing with fire basically, right? So you could hurt yourselves fairly quickly. So that's why when people put this frameworks together, for example, your iPad talked about putting a trust framework together for that, so in which you left your enterprise developers use the whatever the backend you want to use, LLM, foundational model, charge a bit, what have you, but you use it in a controlled way, right? So by putting those frameworks in, you can control to an extent, but it's impossible to stop them from using it because if you put a corporate policies, put a firewall block, they're going to use it using their phone or home computer, they're going to use it. So you got to figure out a way to use it that benefits you rather than hurts you. So you got to put, yeah, responsible. Yeah. All right, final thoughts. Dave, forward six, the cubes, fifth year covering? Wow. Yeah, fifth year, there's COVID in there. Yeah, yeah, yeah. What are your final thoughts on what you saw and heard the last couple of days? You know, every COVID, every COVID, every forward I've been to was better and better and better. You know, it's the Amazon term raising the bar. You guys always raise the bar. This was the best one I've ever been to. Venue, eh, I wasn't crazy about the venue, but because what's happened is UI path keeps growing, so the venue has to grow, and so they're in the tweener venues now. You know, eventually they'll blow through that, but the other thing that strikes me about this conference is it's customer-led. Yes. You know, the content from the customers is really strong. That's a hallmark of a company that's got legs. You saw that with ServiceNow, you saw that with Tableau in the early days, Tableau couldn't remain independent. I asked Daniel Danez on our discussion today, you know, what do you think? What do you have to do to remain independent? He said, I want to stay independent, and we're doing the things necessary to do that. Do so. Independent companies drive innovation. When M&A happens, innovation dies. It just does, as I hate to say it, but it's true. So we like independent companies like UI path, so we're excited about that. And I think this whole notion from point product to platform has clearly taken place. Would you agree? That I agree, so. That puts them in a strong position. Now, having said all that, they got to keep moving fast. They got to move faster than Microsoft. They got to figure out, okay, what are LLMs going to do to the low end of the business? There's a lot of things you could do with ChatGPT that you were doing with RPA before. So they were smart to platformize and move on. So all that was in full display here. And, yeah, it's a fun show. I have always enjoyed this. Yeah, Andy, final thoughts, things that surprised you and impressed you. No problem. First of all, thanks for having me. Thanks for joining us. Yeah, I agree with the platform play. And there are some flashes of innovation angle that they thrown out there, which is somewhat ahead of the market or the other people are not thinking about, right? Again, all of the news, most of the news announcement that came in, it's not in production yet, right? I mean, we are looking at early 2024 is what they told me for most of the functions that's in there. So when it comes in, we got to really see and how it's going to come in. These are all first initial announcements. They're going to shape it. And I'm going to do a deeper briefing with them to see what works, what doesn't work. But from what I've seen, they're trying to pivot from fighting at the same level of low end process automation into a more of a value proposition level. That's a different story to tell, a different budget to attack, different competitors. They could win it. And we didn't talk about iPads, but I actually think their iPad strategy has a good shot because they're not trying to compete with MuleSoft. They're trying to say, hey, we have a new incremental path, basically an ACV path for our customers. So we'll see, I mean, why introduce another iPads if I can get it from, I mean, they're big enough that they may be able to pull that off. In the follow-up conversations, that's one of the discussions we had. You buy it or no? An iPad. So I buy it, the name, I don't buy it because it sounds like as if they're competing in the MuleSofts of the world. But the concept, what they're trying to do of API integration, more of API force orchestration, the workflow and that comes with that and then security and other things to put it in there. And that's where the AI orchestration comes in play as well. And that's logical. Gotta buy it. Yeah, that makes sense. And that's where I think they could have a differentiator. If they execute, I think they'll do well there. That's the point. You know, they have to and they do. And I mean, they have a capable team that they hired all the last year or two and I see them going places. Yeah, this company is executed on product. That's been pretty clear. They took on a big, they took a big bite out of the, that platform play was not trivial. To say we're going from here to here, we're going to make some acquisitions and we're going to integrate, that's a bold move. A space to watch for sure. Thank you so much, all of you for covering this the last couple of days, providing your insights. Definitely a space we're going to keep our eyes on. We appreciate that. We also appreciate you watching theCUBE. There you are. Hi. We've appreciated your viewership over the last couple of days of theCUBE covering UI path forward six from Las Vegas for Dave Vellante, Andy Turai and Rebecca Knight. Lisa Martin, see you next time.