Red Hat Summit 2026 | Ignacio Riesgo, Red Hat, & Jason McGee, IBM

Clips
More from Red Hat Summit 2026

Ignacio Riesgo

Senior Director, Developer Advocacy, IBM & Red Hat Application Development

IBM

Jason McGee

IBM Fellow, CTO IBM Cloud, GM Cloud Platform

IBM

play_circle_outline Scaling Enterprise Agentic AI: From Pilot to Production — Deployment, Infrastructure, Security, and Resiliency

play_circle_outline IBM's Full-Stack Resurgence: GPUs, Accelerators, and Cloud-to-Mainframe Fusion for Secure Enterprise Workloads

play_circle_outline Securing AI Agents: A Data Strategy for Unstructured Institutional Sources, Access Controls, and Evaluation

play_circle_outline Developer role evolving: from coding to intent, agent design, guardrails, and understanding code

Info

Ignacio Riesgo, Red Hat, & Jason McGee, IBM

Ignacio Riesgo

Senior Director, Developer Advocacy, IBM & Red Hat Application Development IBM

Jason McGee

IBM Fellow, CTO IBM Cloud, GM Cloud Platform IBM

search

Rebecca Knight

>> Good afternoon everyone and welcome back to Atlanta where we are theCUBE and we are live at the Red Hat Summit at the Georgia World Congress Center. I'm your host, Rebecca Knight alongside Rob Strechay, and we're going to talk about something that doesn't get enough airtime on this program, which is how organizations can accelerate enterprise AI.

Rob Strechay

>> Yeah. I don't think we've mentioned AI all week long.

Rebecca Knight

>> It's really true. Yeah.

Rob Strechay

>> First time for everything. So, let's dive in here.

Rebecca Knight

>> All right. Well, I'd like to welcome our next guest, Jason McGee, CTO IBM Cloud and GM Cloud Platform. Welcome back, Jason.

Jason McGee

>> Thanks for having me.

Rebecca Knight

>> And Ignacio Riesgo, Senior Director, Developer Advocacy, IBM and Red Hat Application Development. Welcome back to both of you.

Ignacio Riesgo

>> Thank you so much.

Rebecca Knight

>> So let's start with the problem because many organizations are in fact struggling with the transition of AI experimentation and pilot to actually scaling it across their organizations. What do you both see as the real challenges that are throwing them off track? I'll start with you, Jason, and then you, Ignacio.

Jason McGee

>> Sure. I think you're right. I think there's been a big change in the last year as people are definitely doing AI and have probably already deployed some of it, but how do I do it at scale? And I think they're struggling with their strategies around not only what problems to solve, but where should they run, what infrastructure do I need to run it on? And I think we've also transitioned into agentic as the primary pattern that's getting deployed, which is putting more demands on existing backend systems, on traditional computing things like security and resiliency. And so working through that kind of complexity of how do I take an idea and deploy it to my whole organization and make it secure in scale is I think what they're kind of struggling through at the moment.

Rebecca Knight

>> Ignacio?

Ignacio Riesgo

>> Probably it depends on who you ask. If you ask a developer, they will tell you, "Well, my problems are I don't know what to use. If I use Java or if I use Python? If you ask with a CIO. I don't know which model and is secure the code that is coming from that model." If you ask to the DevOps, what do I need to know about agents and how agents are going to talk with the rest of the apps?" So right now, if you think about one year ago we were talking about modernization as one of the critical areas. This year we are completely changing the pace. We are talking about agents, we are talking about LLMs. So the conversation has evolved and now the level of complexity is in another level.

Rob Strechay

>> So let's kind of break it down a little bit and see AI has really come to development teams in mass and I know more broadly they're embracing it and then you have Bob and I got to meet Bob, IBM Bob at TechXchange last year and when it was first announced and all of that, how are people realizing the power of AI and how are they embracing it to really move things forward?

Ignacio Riesgo

>> Well, it has been a year. Okay. For giving you a bit of context. We started with Bob about a year ago internally. We moved from 100 developers using Bob and now we are having 80,000 developers using Bob. And we measure also the productivity and they are getting about 45% more productivity on their own tasks. So the number one single customer has been IBM. You can think about the best use case. So now Bob is a reality. We have IBM Bob there at the booth. A lot of people are taking now hands-on activities. We also ... Remember Bob officially just launched one week ago. So in May is when we make the public announcement. So it's again an AI code assistant that is helping not only IBM but also now you can use, and we will talk about Bob later on in detail, but it's giving you also the autonomy that Bob is going to pick what is the right LLM depending on the question that you ask. So it has been a year and looking forward for the adoption of Bob in customers.

Rebecca Knight

>> Jason, AI sovereignty has become a big part of the conversations in boardrooms and even what we're talking about here on theCUBE. From your perspective at IBM Cloud, what is driving the conversation and how is it different from conversations that we were talking about in terms of data residency just a few years ago?

Jason McGee

>> Yeah, it's really amped up in the last year or two. Obviously a lot of geopolitics that are driving some of that and of course the adoption of cloud overall I think has like as that's become more embedded in people's daily lives, then they are thinking about these kind of sovereign things. I think from our perspective, the most important thing as customers are addressing sovereign is do you have a technology platform that allows you to flexibly decide where you need to run across different cloud providers, across on prem environments, within a country boundary, with a locally owned entity that meets your sovereign requirements. That's why I think some of the things we're talking about here at the Red Hat Summit are so important. Whether Linux or OpenShift or the AI work with Red Hat AI, those all provide the foundations to give customers the ability to deal with these new sovereign requirements to have more control over their destiny and run that wherever they need to. And like your Bob example of even model selection, like when you think about AI, what models do I want to use? Do I use open models or frontier models? You need a platform that gives you the flexibility to make those choices depending on what problem you're trying to solve.

Rob Strechay

>> So we're here at Summit and, again, trying to help people connect the dots. What is it with IBM's infrastructure and Red Hat and how is that really helping accelerate organizations' AI journeys securely?

Jason McGee

>> Yeah, I think it's so important. It's fascinating that infrastructure has kind of reemerged as such a critical element of being successful. And we all think of that of course in things like GPUs and accelerators, because those are a key part of how we're both building and using AI, but we're also putting new demands on traditional compute infrastructure as well. And I think many people have talked about how even the ratios between GPUs and CPUs are shifting very quickly from maybe eight GPUs for one CPU to one for one and that's being driven by agents making lots of API calls and tool calls and driving backend systems and AI is kind of enabling those agents to do the work and all of that is pushing on infrastructure. At IBM, whether it's in our cloud, with IBM Cloud or on premise with mainframes or power or our fusion portfolio we are trying to provide a full stack that allows you to run all those workloads wherever you need them and maintain security and compliance and resiliency and real enterprise characteristics. Because I think that's the difference between a fun pilot and an impactful change to an organization. Can you deploy this for real with my whole team in a real environment?

Rebecca Knight

>> Ignacio, how does that show up in terms of what the developer is working on?

Ignacio Riesgo

>> Well, if you think about where we are now, what an enterprise needs is please give me everything together. So give me the developer tools that I need, the run times so I can continue doing my development, give me also the AI code tools like Bob. Give me, everything that Red Hat provides. If you think about you provide the Linux operative system, RHEL, you provide with the platform, OpenShift, you provide with Ansible. So you put all that together and you say, "And all this is coming from the same company." So you don't need to build your own thing and I just want something that works is reliable, it's giving me all the tools that we need. And by the way, it's compliant and the number one topic from yesterday keynote is sovereignty. Okay So you feed all those boxes from developers to operations to the CIO and you say, "Yeah, we are in a good position."

Rob Strechay

>> So when you look at AI, AI is fed by data and a lot of it is just right data, right place, bringing the AI to the data and the data to the AI and all of that good stuff. What should organizations be really prioritizing at this moment when they think about kind of that data strategy that they're looking at?

Jason McGee

>> I think a couple of things. I mean, one is the data you have is diverse. You can't really make simplistic decisions about data, where it lives, how you prepare it for AI. So, the data on one hand is driving the strategy around infrastructure usage and where AI lives. So like you have to kind of combine all the threads. You can't view data as a separate silo. It's part of the core strategy of an AI based system and you got to have to work through that. I think the second aspect, which is maybe a little bit of tangent, but really interesting to me is that agentic AI systems mean a lot of our normal interactions have to be encoded in a way that agents can get to them. So if you think about phone conversations and Teams meetings and Slack channels and all the ways we communicate, like we need to turn that stuff into data that agents can participate in as well. So we're actually creating a whole new wave of additional data sources that'll enable agents to be parts of our team and interact with us.

Ignacio Riesgo

>> And if you let me, that is, returning to the topic of Bob, if you think about data, there is always data that is visible and there is a lot of data, especially in enterprises, that is . What are those dependencies that if you touch this line of code, that dependency is going to be impacted. All that information is what really Bob is trying to address. That is, we understand not only the inner loop, but also the outer loop, we also understand the production and we are able to connect those pieces through IBM Bob. That's when using these type of tools is addressing both the whole inner loop and outer loop. That's what we call the SDLC, software development life cycle, that is tools that understand your whole life cycle, not only the data, but from the data up to the production.

Rebecca Knight

>> Do you foresee any hurdles or challenges with letting agents in on that unstructured data you're talking about in terms of letting them into our meetings and seeing the transcripts of, and maybe misreading the context of something? What are some of the things that organizations are worried about and how are you assuaging their worries or can you assuage their worries yet?

Jason McGee

>> Yeah, they're absolutely worrying about it, as they should.

Ignacio Riesgo

>> Yeah.

Jason McGee

>> I mean, that information is the institutional knowledge of your company or your organization, it's critical stuff. So there's the kind of maybe basic part of it, which is like, how do you secure and provide access to it? And that drives a lot of the sovereign discussion about where people want to put it and who has access to it and can I be turned off and things like that. I think that aspect of it is like incredibly important. I also think there's a dimension of how do you measure effectiveness of AI, get the right feedback loops in place, get the right evaluation in place so you can kind of check and put like objective reality behind, are the agents accomplishing what I want? Are they doing things in the right way? And I think a lot of organizations are not necessarily thinking deeply enough about feedback loops and evaluation loops that they deploy this AI on a set of data and are they getting what they want out of it?

Rob Strechay

>> Yeah. I mean, I think I look at it as there's also how do you bring together what was the traditional infrastructure and what is the future infrastructure? You guys had an announcement with OpenShift Virt running on IBM Cloud. Want to kind of help us understand how that really helps also bring that together in that way?

Jason McGee

>> Yeah, absolutely. So we announced this week a fully managed OpenShift Virtualization service on IBM Cloud built on Red Hat's OpenShift Virtualization technology. I think a great example of like providing a infrastructure foundation that takes the existing landscape of IT and brings it into a new architecture closer to containers and Kubernetes, which by the way is where all the new AI stuff is getting built, where all the agents are running, those things are all tending to get built on modern cloud native technology platforms. And so by bringing those two things together, we're able to connect more easily the data, the AI systems to the traditional systems and give a simpler model, frankly, also for the agents to use to help you operate that environment, consistent patterns and tools and techniques. So I think it's like all part of this bigger landscape of like, what does actually future IT look like? And AI and accelerators are just going to be part of the fabric of how we solve problems. So the whole landscape has to be brought together.

Rob Strechay

>> And that must help the developers who are looking at it and going, "Hey, it's common tooling," and really brings together and ties into what you guys are doing with Bob and everything else as well. So again, that common platform.

Ignacio Riesgo

>> Yeah. If you think about the role of developer is changing dramatically. If you think about in the past, developers create lines of code and they know exactly what was going to happen. Now with these new tools, first of all you don't need to create anymore that line of code because the AI code assistant is going to create probably faster and better than you. What you need to do is understand that line of code, ensure that the line of code is secure and how you are going to use. So that's why we are thinking that the change is developed by intent that is what is the final objective that you are trying to obtain with that application. So the role of the developer and what the developer needs to learn about is changing dramatically compared with last year. We see an opportunity obviously for those developers that are open and eager to learn about these new things. So instead of developing an app, develop an agent and what that agent is going to be and what are the guardrails that you are going to put into that agent. So again, it's a fast pace, but it's an opportunity for those developers that are eager to learn those, plenty do.

Rob Strechay

>> Yeah. Agreed.

Rebecca Knight

>> So as we are wrapping up here I want to hear both of your personal calls to action for enterprise leaders who are hopefully continuing their AI journeys this year and beyond and what is the one piece of advice you would give them? Ignacio, I'll start with you.

Ignacio Riesgo

>> One advice, I have more than one. Number one, move from experimentation to production, pick one. Pick whatever you want, but move. Okay, stop thinking, no, I am going to try this. No, put something in production. Number two suggestion, data. We say garbage in, garbage out, ensure that the data quality that you have is ... And third, learn those new tools. Okay. Stop thinking, "No, but I know how to code. I know how I do my job." Sorry, the world is changing. There are new tools available for you, start using. And also start comparing because yes, it's true that we put our own AI code assistant that is IBM Bob, but there are also others in the market. So you need to start understanding which to use for what task. Those are my three.

Rebecca Knight

>> Okay. Jason?

Jason McGee

>> I'm going to do more than one as well.

Rebecca Knight

>> All right.

Jason McGee

>> First, I would say really think about your processes and how you can invite agents into that process. How do you do work? How do people communicate? Where do you store the knowledge in your organization and are you explicitly thinking about agents being part of your team to do that? Secondly, you look at the transformation that's happening in spaces like Bob, that transformation's going to happen in all the other domains inside companies. In infrastructure, you have deployment and operation of production infrastructure and applications. Those are going to be handed over to a set of agents that understand that domain as well. So think about how all the other pieces work, how agents are going to influence that and ensure that you have a strategy around your infrastructure and your IT design that's going to support the massive expansion of computation that's going to be needed to make that world real.

Rebecca Knight

>> Excellent. Great advice. Ignacio, Jason, thank you both so much for coming on the show.

Ignacio Riesgo

>> A pleasure as always.

Jason McGee

>> Thanks for having us.

Rob Strechay

>> Thank you.

Ignacio Riesgo

>> Thank you.

Rebecca Knight

>> I'm Rebecca Knight, for Rob Strechay, stay tuned for more of theCUBE's live coverage of the Red Hot Summit. You're watching theCUBE, the leader in enterprise tech news and analysis.

Ignacio Riesgo, Red Hat, & Jason McGee, IBM

search

Rebecca Knight

>> Good afternoon everyone and welcome back to Atlanta where we are theCUBE and we are live at the Red Hat Summit at the Georgia World Congress Center. I'm your host, Rebecca Knight alongside Rob Strechay, and we're going to talk about something that doesn't get enough airtime on this program, which is how organizations can accelerate enterprise AI.

Rob Strechay

>> Yeah. I don't think we've mentioned AI all week long.

Rebecca Knight

>> It's really true. Yeah.

Rob Strechay

>> First time for everything. So, let's dive in here.

Rebecca Knight

>> All right. Well, I'd like to welcome our next guest, Jason McGee, CTO IBM Cloud and GM Cloud Platform. Welcome back, Jason.

Jason McGee

>> Thanks for having me.

Rebecca Knight

>> And Ignacio Riesgo, Senior Director, Developer Advocacy, IBM and Red Hat Application Development. Welcome back to both of you.

Ignacio Riesgo

>> Thank you so much.

Rebecca Knight

>> So let's start with the problem because many organizations are in fact struggling with the transition of AI experimentation and pilot to actually scaling it across their organizations. What do you both see as the real challenges that are throwing them off track? I'll start with you, Jason, and then you, Ignacio.

Jason McGee

>> Sure. I think you're right. I think there's been a big change in the last year as people are definitely doing AI and have probably already deployed some of it, but how do I do it at scale? And I think they're struggling with their strategies around not only what problems to solve, but where should they run, what infrastructure do I need to run it on? And I think we've also transitioned into agentic as the primary pattern that's getting deployed, which is putting more demands on existing backend systems, on traditional computing things like security and resiliency. And so working through that kind of complexity of how do I take an idea and deploy it to my whole organization and make it secure in scale is I think what they're kind of struggling through at the moment.

Rebecca Knight

>> Ignacio?

Ignacio Riesgo

>> Probably it depends on who you ask. If you ask a developer, they will tell you, "Well, my problems are I don't know what to use. If I use Java or if I use Python? If you ask with a CIO. I don't know which model and is secure the code that is coming from that model." If you ask to the DevOps, what do I need to know about agents and how agents are going to talk with the rest of the apps?" So right now, if you think about one year ago we were talking about modernization as one of the critical areas. This year we are completely changing the pace. We are talking about agents, we are talking about LLMs. So the conversation has evolved and now the level of complexity is in another level.

Rob Strechay

>> So let's kind of break it down a little bit and see AI has really come to development teams in mass and I know more broadly they're embracing it and then you have Bob and I got to meet Bob, IBM Bob at TechXchange last year and when it was first announced and all of that, how are people realizing the power of AI and how are they embracing it to really move things forward?

Ignacio Riesgo

>> Well, it has been a year. Okay. For giving you a bit of context. We started with Bob about a year ago internally. We moved from 100 developers using Bob and now we are having 80,000 developers using Bob. And we measure also the productivity and they are getting about 45% more productivity on their own tasks. So the number one single customer has been IBM. You can think about the best use case. So now Bob is a reality. We have IBM Bob there at the booth. A lot of people are taking now hands-on activities. We also ... Remember Bob officially just launched one week ago. So in May is when we make the public announcement. So it's again an AI code assistant that is helping not only IBM but also now you can use, and we will talk about Bob later on in detail, but it's giving you also the autonomy that Bob is going to pick what is the right LLM depending on the question that you ask. So it has been a year and looking forward for the adoption of Bob in customers.

Rebecca Knight

>> Jason, AI sovereignty has become a big part of the conversations in boardrooms and even what we're talking about here on theCUBE. From your perspective at IBM Cloud, what is driving the conversation and how is it different from conversations that we were talking about in terms of data residency just a few years ago?

Jason McGee

>> Yeah, it's really amped up in the last year or two. Obviously a lot of geopolitics that are driving some of that and of course the adoption of cloud overall I think has like as that's become more embedded in people's daily lives, then they are thinking about these kind of sovereign things. I think from our perspective, the most important thing as customers are addressing sovereign is do you have a technology platform that allows you to flexibly decide where you need to run across different cloud providers, across on prem environments, within a country boundary, with a locally owned entity that meets your sovereign requirements. That's why I think some of the things we're talking about here at the Red Hat Summit are so important. Whether Linux or OpenShift or the AI work with Red Hat AI, those all provide the foundations to give customers the ability to deal with these new sovereign requirements to have more control over their destiny and run that wherever they need to. And like your Bob example of even model selection, like when you think about AI, what models do I want to use? Do I use open models or frontier models? You need a platform that gives you the flexibility to make those choices depending on what problem you're trying to solve.

Rob Strechay

>> So we're here at Summit and, again, trying to help people connect the dots. What is it with IBM's infrastructure and Red Hat and how is that really helping accelerate organizations' AI journeys securely?

Jason McGee

>> Yeah, I think it's so important. It's fascinating that infrastructure has kind of reemerged as such a critical element of being successful. And we all think of that of course in things like GPUs and accelerators, because those are a key part of how we're both building and using AI, but we're also putting new demands on traditional compute infrastructure as well. And I think many people have talked about how even the ratios between GPUs and CPUs are shifting very quickly from maybe eight GPUs for one CPU to one for one and that's being driven by agents making lots of API calls and tool calls and driving backend systems and AI is kind of enabling those agents to do the work and all of that is pushing on infrastructure. At IBM, whether it's in our cloud, with IBM Cloud or on premise with mainframes or power or our fusion portfolio we are trying to provide a full stack that allows you to run all those workloads wherever you need them and maintain security and compliance and resiliency and real enterprise characteristics. Because I think that's the difference between a fun pilot and an impactful change to an organization. Can you deploy this for real with my whole team in a real environment?

Rebecca Knight

>> Ignacio, how does that show up in terms of what the developer is working on?

Ignacio Riesgo

>> Well, if you think about where we are now, what an enterprise needs is please give me everything together. So give me the developer tools that I need, the run times so I can continue doing my development, give me also the AI code tools like Bob. Give me, everything that Red Hat provides. If you think about you provide the Linux operative system, RHEL, you provide with the platform, OpenShift, you provide with Ansible. So you put all that together and you say, "And all this is coming from the same company." So you don't need to build your own thing and I just want something that works is reliable, it's giving me all the tools that we need. And by the way, it's compliant and the number one topic from yesterday keynote is sovereignty. Okay So you feed all those boxes from developers to operations to the CIO and you say, "Yeah, we are in a good position."

Rob Strechay

>> So when you look at AI, AI is fed by data and a lot of it is just right data, right place, bringing the AI to the data and the data to the AI and all of that good stuff. What should organizations be really prioritizing at this moment when they think about kind of that data strategy that they're looking at?

Jason McGee

>> I think a couple of things. I mean, one is the data you have is diverse. You can't really make simplistic decisions about data, where it lives, how you prepare it for AI. So, the data on one hand is driving the strategy around infrastructure usage and where AI lives. So like you have to kind of combine all the threads. You can't view data as a separate silo. It's part of the core strategy of an AI based system and you got to have to work through that. I think the second aspect, which is maybe a little bit of tangent, but really interesting to me is that agentic AI systems mean a lot of our normal interactions have to be encoded in a way that agents can get to them. So if you think about phone conversations and Teams meetings and Slack channels and all the ways we communicate, like we need to turn that stuff into data that agents can participate in as well. So we're actually creating a whole new wave of additional data sources that'll enable agents to be parts of our team and interact with us.

Ignacio Riesgo

>> And if you let me, that is, returning to the topic of Bob, if you think about data, there is always data that is visible and there is a lot of data, especially in enterprises, that is . What are those dependencies that if you touch this line of code, that dependency is going to be impacted. All that information is what really Bob is trying to address. That is, we understand not only the inner loop, but also the outer loop, we also understand the production and we are able to connect those pieces through IBM Bob. That's when using these type of tools is addressing both the whole inner loop and outer loop. That's what we call the SDLC, software development life cycle, that is tools that understand your whole life cycle, not only the data, but from the data up to the production.

Rebecca Knight

>> Do you foresee any hurdles or challenges with letting agents in on that unstructured data you're talking about in terms of letting them into our meetings and seeing the transcripts of, and maybe misreading the context of something? What are some of the things that organizations are worried about and how are you assuaging their worries or can you assuage their worries yet?

Jason McGee

>> Yeah, they're absolutely worrying about it, as they should.

Ignacio Riesgo

>> Yeah.

Jason McGee

>> I mean, that information is the institutional knowledge of your company or your organization, it's critical stuff. So there's the kind of maybe basic part of it, which is like, how do you secure and provide access to it? And that drives a lot of the sovereign discussion about where people want to put it and who has access to it and can I be turned off and things like that. I think that aspect of it is like incredibly important. I also think there's a dimension of how do you measure effectiveness of AI, get the right feedback loops in place, get the right evaluation in place so you can kind of check and put like objective reality behind, are the agents accomplishing what I want? Are they doing things in the right way? And I think a lot of organizations are not necessarily thinking deeply enough about feedback loops and evaluation loops that they deploy this AI on a set of data and are they getting what they want out of it?

Rob Strechay

>> Yeah. I mean, I think I look at it as there's also how do you bring together what was the traditional infrastructure and what is the future infrastructure? You guys had an announcement with OpenShift Virt running on IBM Cloud. Want to kind of help us understand how that really helps also bring that together in that way?

Jason McGee

>> Yeah, absolutely. So we announced this week a fully managed OpenShift Virtualization service on IBM Cloud built on Red Hat's OpenShift Virtualization technology. I think a great example of like providing a infrastructure foundation that takes the existing landscape of IT and brings it into a new architecture closer to containers and Kubernetes, which by the way is where all the new AI stuff is getting built, where all the agents are running, those things are all tending to get built on modern cloud native technology platforms. And so by bringing those two things together, we're able to connect more easily the data, the AI systems to the traditional systems and give a simpler model, frankly, also for the agents to use to help you operate that environment, consistent patterns and tools and techniques. So I think it's like all part of this bigger landscape of like, what does actually future IT look like? And AI and accelerators are just going to be part of the fabric of how we solve problems. So the whole landscape has to be brought together.

Rob Strechay

>> And that must help the developers who are looking at it and going, "Hey, it's common tooling," and really brings together and ties into what you guys are doing with Bob and everything else as well. So again, that common platform.

Ignacio Riesgo

>> Yeah. If you think about the role of developer is changing dramatically. If you think about in the past, developers create lines of code and they know exactly what was going to happen. Now with these new tools, first of all you don't need to create anymore that line of code because the AI code assistant is going to create probably faster and better than you. What you need to do is understand that line of code, ensure that the line of code is secure and how you are going to use. So that's why we are thinking that the change is developed by intent that is what is the final objective that you are trying to obtain with that application. So the role of the developer and what the developer needs to learn about is changing dramatically compared with last year. We see an opportunity obviously for those developers that are open and eager to learn about these new things. So instead of developing an app, develop an agent and what that agent is going to be and what are the guardrails that you are going to put into that agent. So again, it's a fast pace, but it's an opportunity for those developers that are eager to learn those, plenty do.

Rob Strechay

>> Yeah. Agreed.

Rebecca Knight

>> So as we are wrapping up here I want to hear both of your personal calls to action for enterprise leaders who are hopefully continuing their AI journeys this year and beyond and what is the one piece of advice you would give them? Ignacio, I'll start with you.

Ignacio Riesgo

>> One advice, I have more than one. Number one, move from experimentation to production, pick one. Pick whatever you want, but move. Okay, stop thinking, no, I am going to try this. No, put something in production. Number two suggestion, data. We say garbage in, garbage out, ensure that the data quality that you have is ... And third, learn those new tools. Okay. Stop thinking, "No, but I know how to code. I know how I do my job." Sorry, the world is changing. There are new tools available for you, start using. And also start comparing because yes, it's true that we put our own AI code assistant that is IBM Bob, but there are also others in the market. So you need to start understanding which to use for what task. Those are my three.

Rebecca Knight

>> Okay. Jason?

Jason McGee

>> I'm going to do more than one as well.

Rebecca Knight

>> All right.

Jason McGee

>> First, I would say really think about your processes and how you can invite agents into that process. How do you do work? How do people communicate? Where do you store the knowledge in your organization and are you explicitly thinking about agents being part of your team to do that? Secondly, you look at the transformation that's happening in spaces like Bob, that transformation's going to happen in all the other domains inside companies. In infrastructure, you have deployment and operation of production infrastructure and applications. Those are going to be handed over to a set of agents that understand that domain as well. So think about how all the other pieces work, how agents are going to influence that and ensure that you have a strategy around your infrastructure and your IT design that's going to support the massive expansion of computation that's going to be needed to make that world real.

Rebecca Knight

>> Excellent. Great advice. Ignacio, Jason, thank you both so much for coming on the show.

Ignacio Riesgo

>> A pleasure as always.

Jason McGee

>> Thanks for having us.

Rob Strechay

>> Thank you.

Ignacio Riesgo

>> Thank you.

Rebecca Knight

>> I'm Rebecca Knight, for Rob Strechay, stay tuned for more of theCUBE's live coverage of the Red Hot Summit. You're watching theCUBE, the leader in enterprise tech news and analysis.