AI Agent Builder Summit | Arun Varadarajan, Ascendion

Clips
News
More from AI Agent Builder Summit

Arun Varadarajan

Chief Commercial Officer (CCO)

Ascendion

What to expect at the AI Agent Builder Summit: Join theCUBE April 16

Causal AI is gaining momentum, with seven in 10 enterprise artificial intelligence professionals either using, experimenting with or planning to adopt it by 2026. That surge in interest reflects a growing demand for systems that don’t just respond, but explain: AI that can reason, justify its decisions and build trust with the humans it serves. As companies evolve beyond generative AI’s headline-grabbing outputs toward more functional, dependable AI architectures, agentic AI systems and causal intelligence emerge as essential components of enterprise-grade solutions

AI Agent Builder Summit preview: How agentic AI is transforming software engineering

Ascendion Inc. will be prominently featured at theCUBE Research’s AI Agent Builder Summit, highlighting its innovative approach to agentic AI and intelligent software engineering platforms.Enterprise digital transformation is accelerating, driven by the rapid evolution from AI assistants to more sophisticated AI agents. Businesses are increasingly leveraging agentic AI to modernize legacy systems, enhance operational efficiency, and transform data infrastructure into intelligent digital architectures.“An agent is an entity that can operate independently and autonomously, doing the work for you,” said Arun Varadarajan (pictured, right), chief commercial officer of Ascendion

play_circle_outline Introduction to AI Agent Builder Summit

play_circle_outline Exploring Ascendion's Impact on Enterprise AI and Approach to Digital Engineering Lifecycle

play_circle_outline Transition from AI assisted to agentic models

play_circle_outline Navigating Challenges and Solutions in Enterprise Digital Transformation

play_circle_outline Ascendion AVA+ platform for powering software engineering with AI capabilities

play_circle_outline Four foundational aspects of the platform: transparency, actor-specific studios, APIfication, agent building

Info
Transcript

Arun Varadarajan, Ascendion

Arun Varadarajan

Chief Commercial Officer (CCO) Ascendion

Arun Varadarajan, CCO of Ascendion, joins theCUBE Research’s Scott Hebner to share insights at the AI Agent Builder Summit. Ascendion, a dedicated force in the field of digital engineering, is recognized for its innovative approach to integrating AI into engineering processes, laying the groundwork for the future of autonomous AI agents.

Varadarajan gives a glimpse into Ascendion's innovative "Engineering to the Power of AI" methodology, as he discusses how AI transforms operational efficiency and system modernization. Key takeaways from the discussi... Read more

explore Keep Exploring

What is the upcoming AI Agent Builder Summit dedicated to exploring? add

What method has Ascendion developed over the last five years that involves infusing AI into their engineering processes? add

What is the difference between an agent and an agentic model? add

What were the fundamental issues that clients were facing during the height of the pandemic in the post-digital era according to the speaker? add

What is the core method and principle that Ascendion AVA+ is founded on for powering software engineering? add

What are the different studios or habitats within the platform, and how do they help different roles such as UI/UX designers, product managers, data engineers, developers, and support engineers in their work? add

bolt Powered by CUBE AI

Arun Varadarajan, Ascendion

search

Scott Hebner

>> Hello. Welcome to this preview edition of the upcoming AI Agent Builder Summit, featuring among others, Ascendion, an industry pioneer in helping organizations re-engineer their enterprise for the era of agentic AI. I am Scott Hebner, the principal analyst for AI here at SiliconANGLE Media and theCUBE Research, and we are so excited to be hosting this industry-unique digital summit dedicated to exploring best practices in building AI agents in agentic workflows. We'll also explore how agentic AI can also automate complex transformational processes such as modernizing legacy systems, enhancing operational efficiency, and transforming outdated data systems. This is also key in helping organizations speed their way to creating more modern, highly adaptive and even intelligent digital business architectures. Today we're thrilled to have Arun Varadarajan, the chief commercial officer at Ascendion, joining us today to preview the AI Agent Builder Summit that kicks off on April 16th, and more importantly, to tap into Arun's real-world experience in helping organizations speed their way to ROI and competitive advantage. And so with that, welcome Arun.

Arun Varadarajan

>> Thank you, Scott. Wonderful meeting you. And it's good to be back on theCUBE. The last time I was on theCUBE was 2019, so it's been a while. Good to be back.

Scott Hebner

>> Oh, that has been a while. We're going to have to get you on even more and more, do some of our podcasts too. But I do appreciate-

Arun Varadarajan

>> For sure.

Scott Hebner

>> Yeah, no, I think it's definitely become a digital first world and what we're finding is not as much reading anymore. It's all these video first and this idea of going to a digital TV for the keep track of what's going on in the enterprise tech marketplace. So these are fun to do and I think they're very productive and certainly for the audience that tunes into all these.

Arun Varadarajan

>> For sure. Sure.

Scott Hebner

>> So very much appreciate you being here. So Arun, tell us a little bit about Ascendion.

Arun Varadarajan

>> Oh yeah, so Ascendion is a full lifecycle digital engineering company that helps clients build products and platforms. We are a full lifecycle engineering shop, so we do everything from ideation through to design, through architecture, through to build through to manage, and of course even modernize. We have an interesting method that has developed over the last five years that we've been building the company called Engineering to the Power of AI, where we infuse AI into our engineering process. And I guess that's something you and I are going to explore and discuss more as we go along. We are located in about 40 different locations across the globe. We have eight engineering centers across India, Philippines, Eastern Europe, North America, LATAM, and we serve a plethora of clients who have global footprints. And as I said, the journey started about five years ago and we've been really building the company ever since.

Scott Hebner

>> That's great. I think you're right at the heart of what customers are going to need, businesses as you move forward from where we are today after the generative AI movement and we get into these more intelligent systems and we start applying it to workflows and everything from software development workflows to customer engagement type workflows, I think it's going to get really interesting out there, but definitely an impressive amount of accomplishments in such short amount of time for you guys, and it was very, very impressive by the way, to be named in the HFS Horizon one leader. That's a big accomplishment, especially given the crowd you were evaluated against. And we'll bring up the graphic there. Just tell us a little bit about that. That's a great accomplishment.

Arun Varadarajan

>> Thank you. Thank you. We are super proud of that recognition. This is a vindication of the thesis behind why we started Ascendion. What I failed to mention earlier was the notion of Ascendion was actually birthed in the height of the pandemic and post-digital era. And both those are very important to note. The height of the pandemic, it was tough to find people, tough to find talent, and even if you found talent, you'd never seen them before. They were virtual, situated somewhere in the planet, and it was really difficult for a company of our size to be able to provide confidence to our clients that here's a company that can deliver and execute on projects. So it became a forcing function for us to look at a completely new operating model. And in the operating model, it was very important for us to be able to deliver and give confidence to our clients that we can ensure that we can bring in consistency in how we deliver our work high quality without compromising on speed and velocity and really drive engineering process in a rigorous manner, which is what gave us, when we started thinking about how do we make this happen, we realized that if we used machine learning and AI, and this was back in 2020 when machine learning was the order of the day, we said, if we can use machine learning to help our engineers do engineering faster and better, maybe that's the place to start. But lo behold, machine learning became deep learning, deep learning became generative AI, generative AI became agentic AI and we've been consuming all of this to really create a very new operating paradigm in terms of how engineering gets delivered, managed and operated.

Scott Hebner

>> Yeah, it's been fascinating having been through several of these transformational cycles over the last 30 years. This one is moving at warp speed, the innovation cycles, and I love the graphic you actually have on your website that shows the predictive models, machine learning, and there was three or four years that was the big focus among enterprises. And then over the last, what, two to three years, it's been generative AI and you quickly established leadership in that space. And now we're sort of moving on to the world of AI agents and you start wiring those things together into agentic systems and workflows. And it has really sparked a lot of interest in the marketplace. I mean, everyone's saying it's going to be the topic of 2025, and we're on this transitionary course to move on from generative AI build around it and on top of it. But I have noticed, not confusion per se, but a lot of different opinions on what is really different between an AI assistant that is feeding off a generative AI and LLM model versus an AI agent that is designed to take it to the next level of intelligence. How would you sort of articulate the transition that we're moving from to?

Arun Varadarajan

>> Oh, I mean, we've been actually, like I said, when we started the company and the journey five years ago, we were AI assisted. We use machine learning to look at different testing patterns, coding patterns. How can we improve the quality of our code? How can we do better testing? How can we do better development, better branching strategy? So we've been AI assisted for the longest of time. But I think that the advent of generative AI, and we of course used language models even back then in our platform to help our engineers operate on different aspects, it could be user experience design, it could be building a new API, it could be whatever, or even replatforming a database, for all of that, we used an AI assisted model where we used machine learning, we used deep learning and even generative AI to help our engineers do things faster. Now, think about the notion of an agent. Who is an agent? An agent is an entity that can operate on its own, independently, autonomously, and do the work for you. So AI assisted, it's a human led model where the AI is helping the human operate better. Now, think of the agentic model. Very different. In the agentic model, you as a human are an observer, a validator and more of a purveyor, if you will, of these agents that are doing work for you. It is totally different. You have a lot more of autonomous agents getting work done, some of them in teams. We actually in our platform can create teams of agents that can go and accomplish a task. And you have manager agents that can break down these tasks, assign it to these different agents, and have the agents go and complete the entire task or the entire job, which could be a string of multiple tasks that each of these agents are purposed, designed and operated to do.

Scott Hebner

>> It's a great way to express it, sort of like co-workers, right, where you're sort of managing a whole set of new co-workers, they're helping you get stuff done, they're taking some autonomous actions, they're helping you think through and plan and problem solve. And yeah, it really is taking it to a whole new level of value. And what I really love about what you guys did, and let's bring up another graphic here, I love how you frame the inevitable challenges that all enterprises I think are eventually going to need to address head on. I mean, I think you frame it very clearly and in a hard-hitting way. Can you just talk a little bit about those three and what you've seen with them?

Arun Varadarajan

>> Oh, absolutely. Absolutely, absolutely, Scott. Remember this was the post-digital era, and I call it the post-digital era, the height of the pandemic when Ascendion was born, and if you remember, a lot of the service providers, mea culpa, I was part of the tribe, we made huge promises to our clients on accelerating digital transformation. And a lot of our clients did not really viscerally experience those outcomes that we committed. And I attribute it to three fundamental issues that clients were facing. One was a crisis of trust. How many times have you seen a service provider walk in the door and say, "I've got these fantastic artifacts, I've got these frameworks and I've got this fantastic talent." And when the rubber meets the road, you don't get the talent that was promised to you. You also don't get the so-called artifacts. A lot of it is slideware and vaporware and there is immediate disillusionment. And more importantly, the service provider never provided the transparency to the client in terms of how his or her software was being delivered and cooked. So that was to me, really creating a crisis of trust. Now, crisis of speed is a derivative of the crisis of trust because when clients said, "I want to do more, I need to push the boundaries and accelerate even further," what did service providers do? They said, "Hey, you need a hundred more people that need to show up on Monday morning." Right? Now, you and I know you add more people, it actually slows things down because humans have this tendency of being error-prone, of not following standards, of not following process. And that causes more consternation in terms of how you drive speed and velocity, and it's an antithesis to speed and velocity in many sense. So there was a crisis of speed. People were not able to move fast enough. And the third big thing was, again, a derivative of the first two crises was a crisis of capital. Because of what had happened, service providers were really not incented to help the clients move fast enough from their legacy applications. And every time the client said, we want to move ahead and modernize, the cost of modernizing was way too high. So they remained in legacy, they had a lot of technical debt and a lot of the capital stuck in the old ways. So we said, you know what? If we are going to start a new company, we have to find a way to address each of these crises. And that's why we said rather than having a person dependent operating model, let's bring in a platform based approach to delivering engineering where we can bring in that degree. So today in our platform, which we'll talk more about down the pike, the platform has the ability to provide our clients real time visibility into how their software is cooking in our kitchen. And at the time of the pandemic, that was super important because there was a lot of mistrust. Are those engineers that I'm deploying really doing the work? So right down to the engineer level, our clients can see today in near real time how they are operating on our client software and how they're building new software for them. And of course, when we brought in the platform with machine learning and AI, our fundamental notion was why should cost, speed and quality be trade-offs? And they were traditionally trade-offs. We said we should not have them as trade-offs and we should be able to deliver on all three with no compromise. So that's why we've come in specifically as a company that can really help our clients accelerate their engineering process, build business cases that can help them modernize, that are very attractive and viable and really move them to the new, which is what clients want, unlocking them from these crises that they were facing.

Scott Hebner

>> Yeah, no, I've talked to dozens and dozens and dozens of companies over the last, let's just say three or four or five months about agentic AI, AI agents and those three really resonate strongly with me because it almost captures the aggregate of everything that I heard. It starts with trying to understand what agents are, back to our previous question, and what the value is and how is it different than generative AI and is it a progressive journey from here or is it rip and replace again? And I think that comes to the notion of the capital. This is really added value at a pace and speed you want to go and how do you unleash the capital and the systems you already have to do new things. So that resonated with me in a big time way. I think speed too, and not just speed in the traditional sense, but speed in everything around you is always changing sense. And the faster you can adapt to new things, new challenges, new requirements, that's an element that makes speed go quicker. So that and I think is one of the value props behind the agents. And then I think most importantly of all, as you pointed out before, you're moving from assistants that are more prompt-oriented, that you're kind of guiding to help you do things to an agent that's a little bit more autonomous, helping you plan. No trust, no ROI from that, right? I mean, trust is going to become the currency of innovation and people aren't going to trust the plan. They're not going to trust a decision. They're not going to certainly trust someone going off, in this case, an agent, and doing stuff without you knowing about it necessarily unless there's trust. And you hit on all three of those, which is really, really good because it basically encapsulates what we all need to focus on as everyone starts this journey.

Arun Varadarajan

>> Without a doubt, without a doubt. So if I may just add on to what you said. The crisis of capital is really palpable today. In fact, with the interest rates going up and the cost of capital and the cost of borrowing going up, not only do we have demand pressures which are causing margin compression for many of our clients, there's also a paucity of capital. So clients are really stuck now, "where do I release the money?" So with our agentic model and our platform, what we are saying is there is an opportunity for you to take cost out of your entire life cycle. How do you design, how do you build, how do you engineer, how do you test, how do you deploy, how do you operate? There are opportunities across the entire engineering value chain to use the agentic model and significantly bring down your cost of operations, releasing money for all the innovation you want to do. And as you rightly said, these agents, we have about today in our roster, we have over 1400 agents inside the platform that attack various engineering problems. We have not tried to boil the ocean and build an agentic platform for let's say, clean energy or anything else. Our focus was we're going to build this platform to fundamentally change how software engineering gets done end to end, and we are seeing some tremendous results.

Scott Hebner

>> Yeah, that's awesome because having worked in this industry and particularly in the software group at IBM for most of my time, getting software out is hard and having quality and then adapting it, and the more it becomes agile and it becomes adaptive, and you're leveraging AI not just in the sense of what you've learned in the past, but is able to adapt to what may happen in the future, like you said, it's going to become more trustworthy and higher quality. It's going to speed your outcomes and it's going to allow you to tap into the capital you have and better leverage the systems that are already in place. And I think it is like a game changer that you're trying to put in place here, and-

Arun Varadarajan

>> It is a game changer. Sorry, I didn't mean to cut you off, but it is a true game changer because one of the biggest challenges in software and why software has never really been industrialized, software engineering, is because the way I write a user story is different from the way you write a user story. The way I write for the same piece of logic, the way I write code will be different from the way you write code. Now just imagine if I standardized all of this, because if I have an agent, let's say I have an agent that develops Spring Boot code, I have already designed the agent to follow all of the standards. It follows the standards, it will not muck around and create new constructs. It'll have the controller class, the model class, all of the things that you need to see in good Spring Boot code as an example. Now you and I know if I give one Epic to four different product managers, you'll get five different user stories. Now, the beauty with this is I can bring consistency there. So right through the engineering process, if I bring in that consistency, rigor, the standards, automatically you're going to get a consistent quality product. And more importantly, I don't know if you're aware of this, but nobody will tell you this, clients today are paying for 30 to 35% rework that most service providers do today. Why do they do rework? Because engineer did not understand the user story, engineer did not conform to the quality standards, engineer did not conform to the security standards. Now all that goes away because we are having the agents follow all of this and get it right the first time is what we all learned. That's what I think agentic AI can do that can bring this into what I'm calling the future of lean software engineering, truly mimicking what happened in the manufacturing world.

Scott Hebner

>> That is a fascinating point that I haven't heard yet in the conversations I had, but I think it's absolutely critical. 'Cause you're right, when you're building software, it's sort of a fusion of art and science, and the ones that I feel do the best in producing consistently good software that they're able to adapt are those that extract from the art and the science the best practices and then codify them, or not codify them in the sense of code, but codify them in the sense of this is how we do things. And what it sounds like you're doing is you're taking these hundreds of agents and you're helping to do that. And it creates the consistency and the standards and the framework, which I think is very much needed not only in building agents and agent systems itself, but in the SDLC platform in general. It gives you guidelines and guardrails, but it doesn't really enforce what you're talking about inherently.

Arun Varadarajan

>> You hit upon two very important points, which I'd like to amplify. One of the things that you said is standards. And I know that the guidelines and best practices and standards were not codified. They were in documents. So today what happens is if there is a standard that we identify, you know what, for this kind of performance that I need to see, I need to see a one microsecond screen performance, if I know that the right code that I need to use to deliver that performance is available, that no longer becomes a guideline. It becomes a standard because I will enforce it the next time around any of these requirements come to that agent. So what happens is every best practice becomes codified. If you have a nutty way or a clever way of doing authentication to ensure that there are no malintended folks showing up at your technology door, you can also put that into how the agents are writing code for authentication, which will change your game completely. So what we are saying is gone are the days where we have documents and guidelines that tell you how to do things. It's going to be systemic, it's going to be in the system.

Scott Hebner

>> Yeah, that's awesome. Well, when we get to the summit, we're going to go in much more detail as we talk about how to build agents and how to build agentic systems and what it can do for you and all that. But why don't you give us a little bit of a preview of Ascendion AVA+, that's your platform for powering software engineering.

Arun Varadarajan

>> Oh, for sure. So like I told you, Ascendion AVA+ is founded on the core method and principle that we call as Engineering to the Power of AI. What is Engineering to the Power of AI? What we've been doing over the last four years, more than four years, is looking at all of the work we do across the software lifecycle, whether you're a product manager understanding requirements, doing a lean canvas and trying to figure out what features to build, or if you're somebody who's writing a user story, who's a product owner, or if you're a developer writing code, or if you're a tester writing test scripts or SRE or support engineer, all of the actors, what we've done is we have actually looked at all the work items they do and asked a simple question, can AI do it better? Can an AI agent do it better? So if you think of our platform, our platform basically brought that thinking and methodology and the way we look at our platform is our platform has got four foundational aspects to it. The first is, you remember when we spoke about transparency, that was absolutely uber critical for us. So the top of our platform has this studio which we call as OneView, which gives our clients a near real-time view of how their software is being engineered by us right down to the developer level. Near real-time. So nobody's waiting for a weekly status report. Nobody's looking at these PowerPoint slides that say, this is green and red. You can see real-time how things are operating at a program level, at a project level, at a developer level, team level, whatever level you want. So that's the first part, the transparency. The second part of the platform is we have multiple studios for different actors in the engineering process. So we have a studio for experience engineers who do experience development, like UI/UX designers. We have a studio for product managers, people who are writing specs and epics and stories. We have a studio for data engineers who do data engineering, who build data pipelines, what have you. We have a developer studio for developers who are writing code. We have a modernization studio that allows you to do all kinds of modernization. Legacy modernization is the biggest challenge for many of our clients. Then we have another studio for support engineers. So basically the second part is we have created these studios or habitats where people can come and leverage the agents that we have built to do their work. So that is where we have the AI-assisted, so that's the AI-assisted part of our platform. Then we have what we call as plugins and APIfication of all of our AI services, which means that if I am, let's say a product manager in Jira, I want to consume the AVA services, I can consume it within Jira. If I'm a developer writing Python and let's say Jupyter, which is one of the IDs, within Jupyter, I can consume all of the AVA services like, "Hey, AVA unit test my code." "Hey AVA, complete my code. Here's the logic." "Hey AVA, create documentation." So there are a bunch of these services that I can experience within my habitat. I don't need to go outside. I can be within my development environment. Or a support engineer who's in service now can experience saying, "Hey, I've seen there's this issue coming in. Tell me how to resolve it." So all of that is the plugin and the APIfication of our platform. Then we have the core of our platform. That is where we build agents. We have actually an agent that can help you build an agent so that it's easy for you to build agents. We can build agents on the fly, and this is where we also build all our workflows. And more importantly, this is where we bake our platform so that the platform is customized for the client we serve. What we have realized, Scott, is that tools don't work in this space. You have to tailor the platform for how the client does their work, because different clients have their own engineering methods, they have their own specs, they have their own naming conventions, they have their own coding standards, they have their own enterprise architecture. So if one of my agents is generating, let's say, a sequence diagram, the only the agent can create a sequence diagram is if it can understand the client's enterprise architecture. I was talking about authentication. I have an agent that can write authentication code, but I can't write generic authentication code. I'll have to write code that is compliant with what the client wants. What is the client's entitlement server, for example, et cetera. So what we do in that layer is build these agents, there's an agent's marketplace. There is an agent place where you can manage these agents and bake them for the client. So that's the four-layered approach that we have to our platform. And yeah, we've been having tremendous success in terms of, for example, one of our clients took almost three, four months to convert Perl scripts, and we were able to do it in matter of a week, the same amount of work. So it's dramatic reduction in time to market, dramatic reduction in terms of cost and fundamentally time to value.

Scott Hebner

>> Really great stuff. It's awesome. Well, unfortunately, it's time for us to start to wrap up here, but we will definitely talk in more depth at the actual session at the Agent Builder Summit, and I think we'll get more a little bit into the use cases and a little bit more depth into everything you just sort of previewed. So make sure, everyone, that you join and let's go, we'll go to the next level here. But before we go, Arun, I got just a question out of curiosity just briefly your two cents on. We had a podcast, Paul Nashawaty, who's the app dev analyst part of our team, and one of the things I've been tracking is, actually one of my predictions for AI this year was that software developers are going to become data scientists, meaning that software developers are going to be successful, say five years from now, are going to have data science skills. They're going to have the ability to build AI capabilities. And the data that we showed, there's a 34% compounded growth rate for demand for jobs that is essentially, you're a software developer and you're an AI engineer and a data scientist. A platform like yours just seems like something that will enable software developers to become AI engineers and data scientists 'cause they have the agents to help them, right?

Arun Varadarajan

>> I have a slightly interesting take, which may be taking what you're saying, but taking it to another interesting level in my opinion. So if you ask me what is the future of engineers? Earlier when I was an engineer, my job was to do something. I was either writing code or I was doing design or architecture. I think the future of engineers will be to manage and design systems like we have. So literally I'm saying you will move from becoming, let's say a coding engineer, let's just take that word for a moment, and you will turn into what I call as a systems engineer, because what you will now be doing is designing systems using a platform like us to get work done. So you will just... It's like this. If you say in the old days when you went to a manufacturing plant, I'm talking about 20, 30 years ago before lean manufacturing came into play, there were line workers who were doing all of the work, right? So think of a PCB manufacturing facility. You had people inserting the chips in the line. My job was to insert the RAM chip and some other chip, an ASIC chip maybe. But then once flexible manufacturing came, you no longer were part of the line, but you were outside the line observing the chip insert machine doing its job, fine-tuning that machine to make sure that the line speed was right, the quality was right. So you're going to move, in my opinion, you don't even need to know AI. You need to be aware of AI because AI will do the work for you. I strongly believe that nobody needs to build AI models anymore. We used to build a lot of AI models and went to AVA. Today we use the foundational models. All we are doing is trying to figure out how to leverage these models to get work done. So you need to be AI aware as an engineer, but you need to now move from trying to do everything to having a system orientation to how can you build a software system that can build software for you, right? Software will build software. That's how we see it.

Scott Hebner

>> Yeah, I think that's right on actually, and it kind of intersects with my point of view there that platforms like AVA+ are going to give superpowers to engineers and developers, and they're just going to be able to do a ton more. No matter which way you look at it, the profession's going to get transformed in terms of what you do every day and the escalating amount of value that you can bring in an organization because the platform's like what you have and agentic AI as a broader topic. Awesome. Well, thank you, Arun. I really, really appreciate you taking the time to be here. Definitely looking forward to learning more at the summit. For all of you, please make sure you visit Ascendion.com to learn more. In the meantime, there's some really great papers out there to read on topics such as Achieving Software Excellence, AI-led Modernization, which we will also talk a little bit more about at the summit and just how to create leadership in genAI and agentic-based SDLCs, software development lifecycle platforms, which we got into a little bit here. But also make sure you reach out to Arun on LinkedIn, or you can visit the AI Agent Builder Summit on theCUBE.net's website and get in touch with Arun. Obviously, you'll see him at the summit, so make sure you're there on April 16th when we start the live broadcast, and we'll be exploring with Ascendion and many others, the key success factors in building AI agents and agentic systems. So the other key thing which Arun hit on was trust. The theme of the summit will be Anchor and Achieving Trust in AI Outcomes. Trust is now the currency of innovation. No trust, no ROI. So visit theCUBE.net, add the summit to your calendar, and we'll see you at the summit. Bye for now.