Exploring the Advancements and Challenges in AI Agent Deployment
John Nay, founder and Chief Executive Officer of Norm Ai, joins theCUBE's special presentation with NYSE Wired, focusing on the upcoming Artificial Intelligence Agent Conference 2025. Hosted by John Furrier, co-founder and co-Chief Executive Officer of SiliconANGLE Media, this insightful discussion covers the pivotal developments in AI infrastructure and the regulatory complexities faced by enterprises.
In this episode, Nay shares their expertise in regulatory AI infrastructure, particularly as it pertains to AI agent deployment in highly regulated sectors. The conversation, hosted by Furrier, delves into the evolving landscape of AI technology, compliance challenges, and the strategic initiatives underway at Norm Ai to address the pressing issues surrounding AI deployment. The discussion provides valuable insights for both technology and policy influencers.
Key takeaways from the discussion include the emphasis on the need for dynamic, real-time compliance frameworks that align with regulatory standards, as emphasized by Nay. Furthermore, the episode highlights how enterprises can leverage existing compliance structures to integrate AI technologies more effectively, offering a glimpse into the future of AI agent scalability and regulation. The conversation underscores the importance of bridging the gap between engineering, policy, and technology for sustainable AI innovation.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Agent Conference. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI Agent Conference.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Register For theCUBE + NYSE Wired: AI Agent Conference
Please fill out the information below. You will recieve an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for theCUBE + NYSE Wired: AI Agent Conference.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: AI Agent Conference. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: AI Agent Conference.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Sign in to gain access to theCUBE + NYSE Wired: AI Agent Conference
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: AI Agent Conference. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Robert Brennan, OpenHands
Exploring the Advancements and Challenges in AI Agent Deployment
John Nay, founder and Chief Executive Officer of Norm Ai, joins theCUBE's special presentation with NYSE Wired, focusing on the upcoming Artificial Intelligence Agent Conference 2025. Hosted by John Furrier, co-founder and co-Chief Executive Officer of SiliconANGLE Media, this insightful discussion covers the pivotal developments in AI infrastructure and the regulatory complexities faced by enterprises.
In this episode, Nay shares their expertise in regulatory AI infrastructure, particularly as it pertains to AI agent deployment in highly regulated sectors. The conversation, hosted by Furrier, delves into the evolving landscape of AI technology, compliance challenges, and the strategic initiatives underway at Norm Ai to address the pressing issues surrounding AI deployment. The discussion provides valuable insights for both technology and policy influencers.
Key takeaways from the discussion include the emphasis on the need for dynamic, real-time compliance frameworks that align with regulatory standards, as emphasized by Nay. Furthermore, the episode highlights how enterprises can leverage existing compliance structures to integrate AI technologies more effectively, offering a glimpse into the future of AI agent scalability and regulation. The conversation underscores the importance of bridging the gap between engineering, policy, and technology for sustainable AI innovation.
play_circle_outlineFocus on software development agents for codebase maintenance and tech debt automation
replyShare Clip
play_circle_outlineKubernetes for Autonomous Dev Agents: Secure, Observable, Event-Triggered Automated PRs for Dependency, Vulnerability, and Framework Migrations
replyShare Clip
play_circle_outlineKubernetes as AI OS/control plane standardizing compute and enabling ecosystem growth
>> Welcome back, everyone, to The Cube here at our NYSE studios. Of course, we are Palo Alto studio connecting Silicon Valley and Wall Street, part of our NYSE wired brand, a Cube original program where we feature the leaders in AI, crypto. Innovation has been the theme. This special interview is really a preview for the agent conference. Simon Chan and the team are holding in May. A great event that's emerged as a small community meetup now full-blown conference. The agentic infrastructure and the demand for developers to build and code agents at an all time high. Robert Brennan's here, co-founder and CEO of OpenHands. Robert, thank you for coming in for this preview. It's been like a series preview. We've interviewed, I think, over a dozen people so far. And I'm super psyched for this agent conference that Simon and the team are putting on. Thanks for coming on.
John Furrier
>> Yeah, thanks for having me. I'm excited to be here.
John Furrier
>> Give a little preview, set the context, obviously the buzz on agents. We've been covering the infrastructure side. It's been talked about. But now with OpenClaw, it's a full tsunami global expansion. The developer community is going nuts. Everyone under the age of 30 in tech is doing it. It's like drugs. It's the summer of love, if you will, for agents. It's blowing me away, actually. It's phenomenal. What are you guys doing? Set the table.
John Furrier
>> Yeah, it is a very exciting new technology, this idea that you can get a large language model to drive an agent in a way that allows it to operate autonomously in its own environment. But it's also scary. These agents, they're hard to predict what they're going to do. And if you give them permission to send emails and modify your calendar and give you access to all your data, there's a lot of risk that it could potentially do something you don't want, either by accident or because you've been compromised in some way. It's actually very easy to do what's called prompt injection, essentially poison the large language model with some instructions telling it to email your boss something that you don't want your boss to see, something like that. So what we do with OpenHands is really provide ways to run these agents in a way that is safe and secure for larger enterprises. So every agent gets run in its own sandbox. You can run as many agents as you want in the cloud simultaneously, and you can essentially dictate what these agents are and are not allowed to do without a human intervening.
John Furrier
>> Explain some of the value propositions and use cases you guys are doing with OpenHands. I love the name. More hands on the project, open source. Give a little teaser to some of the value propositions that you guys have.
John Furrier
>> Yeah. We're specifically focused on software development agents, so agents that are working on a production code base, making changes to that code base to fix a bug, add a new feature. But I would say most importantly, we're focused mainly on tech debt and maintenance-type operations on those code bases. What we find is that for most of the day-to-day feature development, it makes sense for an engineer to be pair programming with an agent locally on their laptop. But with OpenHands, you can run fully automated tasks end to end in the cloud. And that works really well for things that are super automatable, like updating your dependencies in your application, resolving vulnerabilities inside the application, migrating from an old version of a framework to a new version of that framework. All that, keep the lights on type work that goes into keeping your code base healthy.
John Furrier
>> Yeah. And obviously with the cloud native conversion to AI, it's great. In fact, at GTC last week, Jensen Huang, the CEO and founder of Nvidia, he said Kubernetes many times on stage, but also in sessions. Kubernetes is the runtime. It's the operating layer. It's the OS of AI. In fact, KubeCon is happening in Europe this week with The Cube there. I just wrote a post on LinkedIn. And I want to get your reaction to this. I think this is right in line with what we're seeing. I want to get your thoughts. The title is The AI Execution Gap Meets Cloud Native Reality. The promise of cloud native was standard infrastructure unlock developer velocity and increase innovation. AI is not yet on the uptake. So only 7% of daily AI deployment in the Linux foundation has been measuring this. 82% Kubernetes adoption. So you're seeing the substrate. This is huge for DevSecOps as AI natives come in and code. Can you share your thoughts on this? Because I think those numbers are going to increase from seven very fast. Again, OpenClaw, just like catnip for developers. So it's going to be crazy. So what's your thoughts?
John Furrier
>> Yeah, we've been Kubernetes first this whole time. I think your average developer today is still running agents locally on their laptop, which is a good way to experiment and really get your hands dirty and start to learn what these agents are capable of. But if you want to run at any kind of scale and if you want to run securely, you really need to be running them inside of Kubernetes where they can be audited by your organization, where they're getting their own sandbox instead of getting access to all the credentials that are locally available on your laptop, et cetera. So I think Kubernetes is the right place to be running agents. And it's where OpenHands has been running them from day one for any of our cloud or enterprise offerings. Everything is inside of Kubernetes.
John Furrier
>> The Cloud native did all the heavy lifting, all that work, stabilize the platform, check. Now AI comes in. I called it the AIOS Kubernetes layer because it truly is a control plane now, but the game is still the same. I want production workloads. So still a lot more development. I mean, yeah, microservices, orchestration, all the hard work done, but now AI's in there. What's the impact? What do you see coming into the cloud native world as cloud native meets up with AI natives?
John Furrier
>> Yeah. I think two big things to look out for. One is you're going to see agents moving at a much bigger scale. You're going to get agents that aren't just being actively piloted by a single person, but that are running autonomously on a schedule or in response to events happening in your cloud. You could imagine something like you get an error in your Datadog logs that immediately triggers an agent to go investigate that error and open a poll request that solves the issue. That's a workflow we see a lot with OpenHands. I think the other thing to look out for is that as those agents become more and more proliferated, there's going to be a much bigger security footprint here. There's going to be a much bigger need to secure these agents, understand what they're doing, audit their behavior, and track them live. So getting some more observability into what these agents are doing, how they're running, how efficient they're being, how safe they're being is going to be super important.
John Furrier
>> Well, you guys been Kubernetes first, of course. You can't really have a good cloud native environment without great containers. The word containers almost got a renaissance again. I feel like there's an act two of containers coming because the catch that I see as an issue, almost seen this... We've seen the movie before. You can't run AI on Kubernetes if it's not optimized. So this is a huge issue. It's almost a replay of the cloud adoption, lift and shift behaviors, poor cost controls, fragmented tooling, lack of standardization at the highest levels. So talk through that redo. What's your observation? What should people be thinking about? Because again, Kubernetes is going to get a big lift off this. It's boring. It's almost like TCPIP. In a way, it's just good standard. Everyone's using it, but now you have containers taking on a much more active role. And so you have this interplay between agentic infrastructure and cloud native. What are some of the thoughts on that?
John Furrier
>> Yeah, I think it's a good metaphor. I think in the early days of cloud, we saw a lot of proprietary solutions proliferating and folks were willing to adopt all of AWS's internal APIs say for managing compute. But then Kubernetes emerged and really standardized the playing field and said, "Hey, there's going to be this lingual Franca for dealing with cloud compute." And the industry immediately jumped on that and everybody had to do a bunch of work to switch over to Kubernetes, but then it became a standard way of running compute in the cloud. And the ecosystem just blossomed as a result. We got a bunch of standard security controls. We got ways to manage containers, manage their cost, manage their access to resources, manage their security profile. So it was a really great boost once we standardized on an open source standard for running cloud compute. I think you're seeing the same thing happen with agents where right now there's a proliferation of proprietary standards for how do you define and run agents. With OpenHands, we're really the leading open source platform for agents here. So we are trying to create that Kubernetes for agents. We're increasingly the open source standard for how you define and run agents at scale. And you can actually see that as new models get announced today. They're often running OpenHands under the hood in order to determine how good those models are at solving coding tasks.
John Furrier
>> I love the closed versus open. That's awesome point. It also brings into lock-in versus portability because when you have open, you have everything out in the open. Of course, you have open models, but that's why open source is thriving, but portability becomes a big thing. How does that impact the agent side of things as we look at this next wave?
John Furrier
>> Yeah, it's super important. I think if you're going to be building all of your developer infrastructure around this new technology, you want to make sure you're building on top of an open source standard. A, so that you don't get locked into any particular vendor, and B, so that you're building on something that could actually become a standard throughout the industry. It's going to be easier to hire for, easier to maintain over time. There's going to be an ecosystem around it. So open is super important here. You also get that interoperability. I think all the big model providers like Anthropic and OpenAI and Google are trying to lock folks into their vertically-integrated ecosystem, but those model providers are leapfrogging each other in terms of capabilities. Every month or two, we see a new model emerge that's better on cost or better on accuracy. And I think having a strategy that is model agnostic that allows you to switch between models without ripping out a bunch of infrastructure is super important. So going with an open source standard like OpenHands is a great way to be future-proof.
John Furrier
>> Well, a lot of tooling discussions, a lot of chaos, which is good. You're seeing growth. Robert, thanks for coming on The Cube. Quick final point, give a preview of what your agenda is at the agent conference. What are some of the conversations you'll be having? What are you looking to hope to accomplish?
John Furrier
>> Yeah, I would say really looking to help the engineering community. A, make this transition into an Agentic world. We're seeing developers start to experiment with agents, especially running locally on their laptop, but it's only the really future-oriented teams that are starting to adopt agents at scale in the cloud. And we're trying to bring folks along for that ride because it's very clear that that's where software engineering is going. And then also trying to do that securely and safely. I think it's a really easy thing to screw up, and so we're trying to help folks navigate that transition.
John Furrier
>> Well, certainly the community's been focused on this, the benefits to the ecosystem, scalable, secure, in production. That's the cloud native playbook come into AI pretty quickly.
John Furrier
>> Yeah. It's an exciting time.
John Furrier
>> Robert, thank you so much for coming on preview of the Agent Conference. Agentconference.com, check it out. Simon Chan and the community growing very rapidly in the AI infrastructure leaders as the world explodes with OpenClaw, a lot more attention on the value, the economics, but the engineering still comes back down to software. You got to code these agents. You could coding these agents with compliance, with governance, all part of the cloud native mindset. We'll see you there in May. I'm John Furrier, your host of The Cube. Thanks for watching.
>> Welcome back, everyone, to The Cube here at our NYSE studios. Of course, we are Palo Alto studio connecting Silicon Valley and Wall Street, part of our NYSE wired brand, a Cube original program where we feature the leaders in AI, crypto. Innovation has been the theme. This special interview is really a preview for the agent conference. Simon Chan and the team are holding in May. A great event that's emerged as a small community meetup now full-blown conference. The agentic infrastructure and the demand for developers to build and code agents at an all time high. Robert Brennan's here, co-founder and CEO of OpenHands. Robert, thank you for coming in for this preview. It's been like a series preview. We've interviewed, I think, over a dozen people so far. And I'm super psyched for this agent conference that Simon and the team are putting on. Thanks for coming on.
John Furrier
>> Yeah, thanks for having me. I'm excited to be here.
John Furrier
>> Give a little preview, set the context, obviously the buzz on agents. We've been covering the infrastructure side. It's been talked about. But now with OpenClaw, it's a full tsunami global expansion. The developer community is going nuts. Everyone under the age of 30 in tech is doing it. It's like drugs. It's the summer of love, if you will, for agents. It's blowing me away, actually. It's phenomenal. What are you guys doing? Set the table.
John Furrier
>> Yeah, it is a very exciting new technology, this idea that you can get a large language model to drive an agent in a way that allows it to operate autonomously in its own environment. But it's also scary. These agents, they're hard to predict what they're going to do. And if you give them permission to send emails and modify your calendar and give you access to all your data, there's a lot of risk that it could potentially do something you don't want, either by accident or because you've been compromised in some way. It's actually very easy to do what's called prompt injection, essentially poison the large language model with some instructions telling it to email your boss something that you don't want your boss to see, something like that. So what we do with OpenHands is really provide ways to run these agents in a way that is safe and secure for larger enterprises. So every agent gets run in its own sandbox. You can run as many agents as you want in the cloud simultaneously, and you can essentially dictate what these agents are and are not allowed to do without a human intervening.
John Furrier
>> Explain some of the value propositions and use cases you guys are doing with OpenHands. I love the name. More hands on the project, open source. Give a little teaser to some of the value propositions that you guys have.
John Furrier
>> Yeah. We're specifically focused on software development agents, so agents that are working on a production code base, making changes to that code base to fix a bug, add a new feature. But I would say most importantly, we're focused mainly on tech debt and maintenance-type operations on those code bases. What we find is that for most of the day-to-day feature development, it makes sense for an engineer to be pair programming with an agent locally on their laptop. But with OpenHands, you can run fully automated tasks end to end in the cloud. And that works really well for things that are super automatable, like updating your dependencies in your application, resolving vulnerabilities inside the application, migrating from an old version of a framework to a new version of that framework. All that, keep the lights on type work that goes into keeping your code base healthy.
John Furrier
>> Yeah. And obviously with the cloud native conversion to AI, it's great. In fact, at GTC last week, Jensen Huang, the CEO and founder of Nvidia, he said Kubernetes many times on stage, but also in sessions. Kubernetes is the runtime. It's the operating layer. It's the OS of AI. In fact, KubeCon is happening in Europe this week with The Cube there. I just wrote a post on LinkedIn. And I want to get your reaction to this. I think this is right in line with what we're seeing. I want to get your thoughts. The title is The AI Execution Gap Meets Cloud Native Reality. The promise of cloud native was standard infrastructure unlock developer velocity and increase innovation. AI is not yet on the uptake. So only 7% of daily AI deployment in the Linux foundation has been measuring this. 82% Kubernetes adoption. So you're seeing the substrate. This is huge for DevSecOps as AI natives come in and code. Can you share your thoughts on this? Because I think those numbers are going to increase from seven very fast. Again, OpenClaw, just like catnip for developers. So it's going to be crazy. So what's your thoughts?
John Furrier
>> Yeah, we've been Kubernetes first this whole time. I think your average developer today is still running agents locally on their laptop, which is a good way to experiment and really get your hands dirty and start to learn what these agents are capable of. But if you want to run at any kind of scale and if you want to run securely, you really need to be running them inside of Kubernetes where they can be audited by your organization, where they're getting their own sandbox instead of getting access to all the credentials that are locally available on your laptop, et cetera. So I think Kubernetes is the right place to be running agents. And it's where OpenHands has been running them from day one for any of our cloud or enterprise offerings. Everything is inside of Kubernetes.
John Furrier
>> The Cloud native did all the heavy lifting, all that work, stabilize the platform, check. Now AI comes in. I called it the AIOS Kubernetes layer because it truly is a control plane now, but the game is still the same. I want production workloads. So still a lot more development. I mean, yeah, microservices, orchestration, all the hard work done, but now AI's in there. What's the impact? What do you see coming into the cloud native world as cloud native meets up with AI natives?
John Furrier
>> Yeah. I think two big things to look out for. One is you're going to see agents moving at a much bigger scale. You're going to get agents that aren't just being actively piloted by a single person, but that are running autonomously on a schedule or in response to events happening in your cloud. You could imagine something like you get an error in your Datadog logs that immediately triggers an agent to go investigate that error and open a poll request that solves the issue. That's a workflow we see a lot with OpenHands. I think the other thing to look out for is that as those agents become more and more proliferated, there's going to be a much bigger security footprint here. There's going to be a much bigger need to secure these agents, understand what they're doing, audit their behavior, and track them live. So getting some more observability into what these agents are doing, how they're running, how efficient they're being, how safe they're being is going to be super important.
John Furrier
>> Well, you guys been Kubernetes first, of course. You can't really have a good cloud native environment without great containers. The word containers almost got a renaissance again. I feel like there's an act two of containers coming because the catch that I see as an issue, almost seen this... We've seen the movie before. You can't run AI on Kubernetes if it's not optimized. So this is a huge issue. It's almost a replay of the cloud adoption, lift and shift behaviors, poor cost controls, fragmented tooling, lack of standardization at the highest levels. So talk through that redo. What's your observation? What should people be thinking about? Because again, Kubernetes is going to get a big lift off this. It's boring. It's almost like TCPIP. In a way, it's just good standard. Everyone's using it, but now you have containers taking on a much more active role. And so you have this interplay between agentic infrastructure and cloud native. What are some of the thoughts on that?
John Furrier
>> Yeah, I think it's a good metaphor. I think in the early days of cloud, we saw a lot of proprietary solutions proliferating and folks were willing to adopt all of AWS's internal APIs say for managing compute. But then Kubernetes emerged and really standardized the playing field and said, "Hey, there's going to be this lingual Franca for dealing with cloud compute." And the industry immediately jumped on that and everybody had to do a bunch of work to switch over to Kubernetes, but then it became a standard way of running compute in the cloud. And the ecosystem just blossomed as a result. We got a bunch of standard security controls. We got ways to manage containers, manage their cost, manage their access to resources, manage their security profile. So it was a really great boost once we standardized on an open source standard for running cloud compute. I think you're seeing the same thing happen with agents where right now there's a proliferation of proprietary standards for how do you define and run agents. With OpenHands, we're really the leading open source platform for agents here. So we are trying to create that Kubernetes for agents. We're increasingly the open source standard for how you define and run agents at scale. And you can actually see that as new models get announced today. They're often running OpenHands under the hood in order to determine how good those models are at solving coding tasks.
John Furrier
>> I love the closed versus open. That's awesome point. It also brings into lock-in versus portability because when you have open, you have everything out in the open. Of course, you have open models, but that's why open source is thriving, but portability becomes a big thing. How does that impact the agent side of things as we look at this next wave?
John Furrier
>> Yeah, it's super important. I think if you're going to be building all of your developer infrastructure around this new technology, you want to make sure you're building on top of an open source standard. A, so that you don't get locked into any particular vendor, and B, so that you're building on something that could actually become a standard throughout the industry. It's going to be easier to hire for, easier to maintain over time. There's going to be an ecosystem around it. So open is super important here. You also get that interoperability. I think all the big model providers like Anthropic and OpenAI and Google are trying to lock folks into their vertically-integrated ecosystem, but those model providers are leapfrogging each other in terms of capabilities. Every month or two, we see a new model emerge that's better on cost or better on accuracy. And I think having a strategy that is model agnostic that allows you to switch between models without ripping out a bunch of infrastructure is super important. So going with an open source standard like OpenHands is a great way to be future-proof.
John Furrier
>> Well, a lot of tooling discussions, a lot of chaos, which is good. You're seeing growth. Robert, thanks for coming on The Cube. Quick final point, give a preview of what your agenda is at the agent conference. What are some of the conversations you'll be having? What are you looking to hope to accomplish?
John Furrier
>> Yeah, I would say really looking to help the engineering community. A, make this transition into an Agentic world. We're seeing developers start to experiment with agents, especially running locally on their laptop, but it's only the really future-oriented teams that are starting to adopt agents at scale in the cloud. And we're trying to bring folks along for that ride because it's very clear that that's where software engineering is going. And then also trying to do that securely and safely. I think it's a really easy thing to screw up, and so we're trying to help folks navigate that transition.
John Furrier
>> Well, certainly the community's been focused on this, the benefits to the ecosystem, scalable, secure, in production. That's the cloud native playbook come into AI pretty quickly.
John Furrier
>> Yeah. It's an exciting time.
John Furrier
>> Robert, thank you so much for coming on preview of the Agent Conference. Agentconference.com, check it out. Simon Chan and the community growing very rapidly in the AI infrastructure leaders as the world explodes with OpenClaw, a lot more attention on the value, the economics, but the engineering still comes back down to software. You got to code these agents. You could coding these agents with compliance, with governance, all part of the cloud native mindset. We'll see you there in May. I'm John Furrier, your host of The Cube. Thanks for watching.
John Furrier