In this theCUBE + NYSE Wired: Mixture of Experts segment from the New York Stock Exchange, theCUBE’s John Furrier sits down with Raj Verma, CEO of SingleStore, to unpack how the intersection of technology and finance is shaping enterprise strategy. Verma shares why SingleStore is “on course” for the public markets, reflects on brand-building through the company’s partnership with golf Hall of Famer Padraig Harrington and connects that ethos to how SingleStore helps organizations fix struggling data “swings.” The discussion zeroes in on what’s next as Wall Street watches the AI infrastructure buildout: after chips and systems, the software and data layers set the pace for value creation.
Verma outlines why enterprises must modernize “brown” data estates into “green” ones to safely bring corporate context, governance and compliance into LLM workflows via RAG – and why commoditized data-at-rest puts the advantage at the query layer that unifies data in motion with data at rest. He predicts agentic AI will gain reasoning capabilities in roughly 18 months, cites industry indicators like Google reporting ~25% of its software now built by AI and argues that high switching costs will give way to disruption as buyers reassess legacy vendors. The conversation closes with concrete momentum: ~33% YoY growth, ARR in the ~$135M range, gross dollar retention ~98%, cloud NDR ~130, ~50% of business now in the cloud, landing ~3 new customers per day, a path to cash-flow breakeven in the next two quarters and a teaser for AI-related announcements in the next two months. Listeners will find notable stats, real-world use cases and forward-looking views on how databases power reliable AI at enterprise scale.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: Mixture of Experts Series. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: Mixture of Experts Series.
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: Mixture of Experts Series
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: Mixture of Experts Series.
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: Mixture of Experts Series. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: Mixture of Experts Series.
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: Mixture of Experts Series
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: Mixture of Experts Series. 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
Jim Zemlin, Linux Foundation & Mazin Gilbert, Agentic AI Foundation
In this theCUBE + NYSE Wired: Mixture of Experts segment from the New York Stock Exchange, theCUBE’s John Furrier sits down with Raj Verma, CEO of SingleStore, to unpack how the intersection of technology and finance is shaping enterprise strategy. Verma shares why SingleStore is “on course” for the public markets, reflects on brand-building through the company’s partnership with golf Hall of Famer Padraig Harrington and connects that ethos to how SingleStore helps organizations fix struggling data “swings.” The discussion zeroes in on what’s next as Wall Street watches the AI infrastructure buildout: after chips and systems, the software and data layers set the pace for value creation.
Verma outlines why enterprises must modernize “brown” data estates into “green” ones to safely bring corporate context, governance and compliance into LLM workflows via RAG – and why commoditized data-at-rest puts the advantage at the query layer that unifies data in motion with data at rest. He predicts agentic AI will gain reasoning capabilities in roughly 18 months, cites industry indicators like Google reporting ~25% of its software now built by AI and argues that high switching costs will give way to disruption as buyers reassess legacy vendors. The conversation closes with concrete momentum: ~33% YoY growth, ARR in the ~$135M range, gross dollar retention ~98%, cloud NDR ~130, ~50% of business now in the cloud, landing ~3 new customers per day, a path to cash-flow breakeven in the next two quarters and a teaser for AI-related announcements in the next two months. Listeners will find notable stats, real-world use cases and forward-looking views on how databases power reliable AI at enterprise scale.
Jim Zemlin, Linux Foundation & Mazin Gilbert, Agentic AI Foundation
Jim Zemlin
Executive DirectorLinux Foundation
Mazin Gilbert
Executive DirectorAgentic AI Foundation
search
John Furrier
>> Welcome back everyone to theCUBE here at the New York Stock Exchange Studios of theCUBE. Of course, we have our Palo Alto studio connecting Silicon Valley and Wall Street. I'm John Furrier, host of theCUBE. This is a preview of what's going on the Agentic Foundation. Of course, the MCP Dev Conference in New York City, the Linux Foundation is happening. A lot's going on in agents and in open source. We have two great guests from the Linux Foundation, Jim Zemlin, CUBE alumni, the CEO of the Linux Foundation and Mazin Gilbert, Executive Director, newly appointed Executive Director of the Agentic AI Foundation. Jim, great to see you. Madison, thanks for coming on theCUBE.
Mazin Gilbert
>> Thank you for having us.
Jim Zemlin
>> Thanks for having us.
John Furrier
>> Best time to be an open source. It's a whole nother level. Everyone's addicted to OpenClaw. You got open source thriving. Agents ramping up on top of Cloud Native. Amazing success with the Linux Foundation. Thanks for doing that, but a lot going on. What's the hard news?
Jim Zemlin
>> Well, so we launched the Agentic AI Foundation in December last year. And in just a few months, the organization has gotten larger than our previously largest organization, the Cloud Native Computing Foundation, home to Kubernetes, with over 170 members today, a ton of momentum. And in that three months, I have been searching for somebody to help me run the organization, and I have finally found him in Mazin here. Mazin is going to be the new executive director of the Agentic AI Foundation, and he has all the right ingredients for leading this organization. One, he has a PhD, which in the world of artificial intelligence, as you know, is a prerequisite. In fact, I think he wrote his thesis for his PhD program on neural networks. But he also has an MBA from the Wharton School. So he not only understands the tech side, which your Palo Alto office can testify to, but also this, the business side from his experience getting his MBA. But even more importantly, he spent his last five years at Google building AI solutions there. So he has all the right pedigrees to lead the fastest growing open source organization in the world. We're super, super happy to have Mazin on board.
John Furrier
>> Mazin, congratulations. I'm super excited that you're here. One, the event is happening in New York City with MCP, which is just a feeding ground for developers and innovation. But this foundation is looking from all the signals. It's going to be rapidly growing. The CNCF became really a key enabler for what we're now seeing. Took a little time to get off the runway because it was standards were growing, it was developing. A lot of hard work was done. But that work in the CNCF has set the table for the AI wave, which is highly robust, rapidly evolving at all levels. You got to pull all your tricks out of the bag on this one. Give us your thoughts and vision for the foundation.
Mazin Gilbert
>> Well, first, thank you very much for the invitation. Thank you, Jim. It's an exciting, amazing time right now. You see this space of agentic AI is moving at really the speed of light. We've never seen that kind of growth before. Companies are coming out of nowhere and coming out with some new protocols and getting adopted. Millions of people are adopting it in days or a week or two. Okay? So we've never seen this before. We're embarking into an era that I call the internet of agents. We've gone from the cloud native to the AI native era to this internet of agents where autonomous agents are going to be collaborating, reasoning. They're going to be taking actions and fulfilling those actions. We may end up with more agents in the world in the next three, five years than even human beings that we have today, which is an exciting and amazing world. The foundation is really important here, because we've seen this playbook before with CNCF. We've seen it with networking as well, with the Linux Foundation Networking. When you go from experimentation to production, you really need a neutral, open standard community that will really define the protocols, the plumbing, the safety, the auditability, the observability to make it enterprise ready. And this is what really this foundation is. And really the vision of the foundation is to create that open standard and make it real with all these participants, 170 companies and counting.
John Furrier
>> Jim and I have talked about this in the past on theCUBE and we've had many conversations at the Linux Foundation events about open source. And remember when I was in college, that was the proprietary software days and it was just open source was an underground movement at that time. So I saw that proprietary wave. A lot of us who are pioneering and open source were there and saw that growth and then everything just magically happened and here we are today. But the key inflection points at all these major ways, whether it's proprietary to open source, mini computers to open networking, standards were super important. If you look at the OSI model with TCP/IP, that was a huge thing. When you look at the internet, DNS, HTTP, certain standards that weren't restrictive, but yet open became a huge deal. I mean, cloud kind of was a standard, but I think the internet and AI kind of have a similar pattern, but at different scales, the need for that enablement.
Mazin Gilbert
>> Yeah.
John Furrier
>> Talk about that, because I think we are in a moment of consequence where at this point in time with AI and the massive innovation, that standards play, specifically standards play a key role.
Jim Zemlin
>> I think the way to think about why is agent technology so big, start there. So for decades, Silicon Valley technology companies built technology tools for you and I to use. You would log in to use a spreadsheet, you would log in to use a SaaS application. People used technology. Agents are very different than that. Agents are technology that uses other tools and technologies. That expands the total addressable market, the impact on society to a much more profound thing. And so if you think of agents as technology that uses other technology, the first question you have to ask is there are millions of different technology products and services out there. How can these agents efficiently connect to all of those different technology products and services? Well, the way you make that efficient is you have a standardized way to allow those agents to connect to that data and those technology services. And that is MCP, which is the core project and core specification at the Agentic AI Foundation. But you also need a whole bunch of other things around that core protocol. You need open source tools to allow people to create a way to deploy agents, say on their desktop. You need other adjacent technology around identity and access management, all the things that Mazin talked about. And that is what we're building at the Agentic AI Foundation.
John Furrier
>> Mazin, what's your vision on this? Because the aperture of participation, we think at theCUBE here, and I think you guys might agree, isn't it just your classic developer. There's a lot of stakeholders. AI is for everyone. It's infused in every single vertical. Vertical apps is the hottest market right now. I think that's just a predictor, a pretext, in my opinion, to massive proliferation of AI. There's a lot of even CFO conversations. Jensen Huang said token budget for your employment contract at GTC, hinting that your performance in your job will be determined by how well you use your tooling, or in this case, agents. Reminds me when sales reps had cell phone budgets, if it wasn't huge, they weren't doing their job calling customers. So you have a similar telecom AI kind of dynamic going on where tokens and for agents become not just a technical thing.
Mazin Gilbert
>> Yeah. So if you think about this, tokens to LLMs is agents are to applications. So when Jensen was talking about token budget, we're going to see a future as we get into this internet of agents is that you have an agent budget. How many agents do you have doing all sort of different type of work for you from all the kind of different verticals? If you look at the members today of the foundation, a lot of them are technology members driving the foundation protocols, the tools. As Jim basically mentioned, the training, you need that solid foundation. Once we get that and once we harmonize that, and that's what we're doing over the next weeks and months, then the verticals become really important. These are the vertical applications in finance and telecommunications and retail, et cetera. So you're going to see a lot more adoption of applications, not just number of downloads of these protocols. Like today for MCP, we have 97 million downloads a month of that protocol. Just imagine that. Okay? It's hard to believe that kind of sheer size. What we are going to go to is not just downloads of protocols. We're going to go into how many of these agents are in production and deployed. And that's really where we want to get to as a foundation, is to go from the experimentation to production to make this real,, actually solving real business problems and driving the return on investment.
John Furrier
>> It's interesting. The cloud was such a disruptive enabler and accelerator too, but the dimensions involved on the business model impact wasn't that complicated. Go to subscription model, pay by the drink, go to subscription. So easy to understand. Agents, not so much. I mean, you mentioned tokens, that's like a headcount. It's like almost a staffing challenge. It has nothing to do with technology. So less of an IT and problem that cloud solved and scaled. Now, agents are a little bit more. So it's interesting to see even some of the entrepreneurial activities are started by people that aren't even coders.
Mazin Gilbert
>> Yes.
John Furrier
>> So you have a whole nother level going here. Do you guys talk about that when you have meetings?
Jim Zemlin
>> I mean, one of the fun things about working on the Agentic AI Foundation is we get to talk to a lot of end users who are deploying agents a lot in proof of concepts initially, some in production. And we get to hear about gaps that they have as they roll that out and what they need in terms of standard methods, in terms of security, in order to make those really successful. Like for example, if you're delegating a bunch of work to an agent, how do we handle identity? Who is working on behalf of what? What identity does that agent have? You talk about cost control, right? So you have all these agents. What's the control plane in order to manage how many tokens each agent is consuming? As an agent manager, which is now a new job description, how do I make sure that I have some way of looking across all these agents and stopping one before it spends my entire token budget for the month?
John Furrier
>> Well, we had FinOps for cloud. I'm sure there'll be an AI version of that.
Jim Zemlin
>> That's right.
John Furrier
>> I mean, Mazin, this is an interesting dynamic. One of the things I was very fascinated with from a historical perspective, looking back at the CNCF and KubeCon and all the adjacency around cloud native was the participation by end users in the foundation. For example, Netflix, Lyft, Intuit. These are companies that traditionally aren't vendors selling stuff or trying to influence the market. They had real code. They were contributing some of the most consequential products in CNCF. If you take that pattern and map it to AI, you're going to have much broader end user participation because they're interested too. Open source means it's open to everybody.
Mazin Gilbert
>> And I think that's what's really exciting about this is that the barrier to entry, to contribute code, to diagnose code, to debug a code has basically lowered down significantly. Many companies who don't, like you said, don't code are now participating. And we are really looking for all these segments of companies. They're the technology companies, which traditionally are the ones who are contributing code. Then you have the large customers, smaller, medium and large customers who are already the end users who are starting to see them also contributing code. But we also care about many other companies who may not be members of the foundation or members of Linux. These are developers. These could be people in universities or colleges or in school and high school who are actually writing code or using agents to write code. And we want to grow that community as well. All of these are really an incredible sort of set of assets for this foundation.
John Furrier
>> Well, we were lucky enough to be there present at creation during pre-CNCF, watch that grow, being a participant with you guys, excited to be part of the Agentic AI Foundation. I'm sure it's going to grow very much faster. So very excited. Thanks for that. Well, we got an event happening in New York. What's the action at the MCP Dev New York activities? Give us a taste of what you're seeing and what you're announcing.
Jim Zemlin
>> We got a sold out house. So we're just off of Times Square and we have all of the core maintainers of the MCP protocol, of the Goose reference framework, of a whole bunch of different open initiatives around agents. But I think what's really interesting, and this is, I think, something important for people who are watching this to understand is the conversation around successful implementation of agentic technology is happening at a real grassroots level. And this event this week is such a great example of that. You have retail companies, Nordstrom, coming in, talking with fellow practitioners, explaining what they're seeing, explaining how they're succeeding, how they're failing. Companies like JPMorgan, you've got financial services companies here talking about agentic success and agentic struggles. And that is really what is so amazing about both this agentic movement and about these kind of events. I mean, it really, and I'm going to date myself here, it hearkens back to those kind of computing user groups that I used to go to as a kid where there was just all this incredible grassroots energy. Everybody is a builder. People are super excited. And that's what's happening here in New York this week.
John Furrier
>> It's interesting you bring that up because I was just talking yesterday on our NYSE Wired program around this new layer of persona that's developing superfast in the agentic infrastructure piece. It looks a lot like DevOps, which then became DevSecOps, which is cloud native, but it's a similar kind of culture, different people. In fact, when I wrote my post covering KubeCon EU, my opening sentence was cloud native, meet AI native. This is synergy between the two. And you talk about the numbers on the downloads, it's mass scale, but the grassroots is still an emerging community. That layer of the agentic infrastructure is propelling the AI native developer. So you have a grassroots movement at scale. I mean, when you look at that, Jim, and historically we say, "How does that happen?" I mean, usually you can't have scale in grassroots. It's either both.
Jim Zemlin
>> Yeah. I think the important thing to remember is even though we're seeing millions of downloads, rapid adoption, I think we're still in the first five minutes of a long game here. And here's why. It's easy to get quick results with agents on very specific tasks, but if you are the CEO of a company and you want to deploy agents at scales, there is a lot of things you need to think about. Security, identity and trust, access control, all of the things that you would need to really hook this up to mission-critical systems. And I think that is where the Agentic AI Foundation is creating real value, creating the standards that allow you to potentially manage a fleet of agents, making sure that security and trust is integrated into all of these openly available tools so that you can hook an agent up, not just to a low risk, maybe customer support system, but to your core financial management system. And you need a lot of different technology to do that, to make sure that private information doesn't leak, to make sure that this stuff doesn't get hacked. There's a lot of things. That is what you're going to see in the coming months coming out of the Agentic AI Foundation is a whole suite of technologies that will be contributed by the different organizations that are participating.
John Furrier
>> Mazin, any thoughts right now as you look at day one in the job? The enthusiasm's high, but the key to success in a lot of these big foundation like CNCF, what's successful was they gave a lot of confidence. You had enthusiasm and confidence in the community. Thoughts on how you see that playing out standards wise or projects? What's your thoughts on that?
Mazin Gilbert
>> We have a rare, unique opportunity and we have the best of the best at the table, the best companies in the world, the biggest companies and the smallest, and the ones that are making the headlines today, they're all the table. The funny part of the dynamics is that those are companies that competed during the day and they are really working furiously together to try to drive those standards and basically these protocols. I think time is critical here. It is the first five minutes, but the next 5 and 10 minutes matters. We have three projects right now as part of the foundation, MCP and we have Goose and AGENTS.md. The agent stack is very rich. And for us to really drive that safety, that trust, that security, there are a lot of opportunities there to bring in new capabilities, to bring in new projects, and we need to bring them fast, because projects are moving very fast right now. And if we don't do that, there will be a lot of fragmentation in the market. And yes, we will get to an exciting agent era, but it will lack standardization and it will be very expensive and that means building at scale, it will be extremely basically slow.
John Furrier
>> I mean, fragmentation encourages proprietary.
Mazin Gilbert
>> That is exactly right. And this is the best time. And I can't tell you we have two, three years. We don't. We have weeks and months. And you're going to see from the foundation over the next weeks and months, we're going to make some really exciting announcements looking at the stack from top to bottom and really coming out and bringing in some very critical projects that's going to help us sort of to really bring agents out at a scale in a secure basically way, including what are the guardrails? And when does a human come in the loop? Because agents are autonomous, but there are times you want a human in the loop. These all have to be standardized.
John Furrier
>> Yeah. I love the hype cycle, because I can tell you when the hype cycle's real because the demand for the long form CUBE content is high. AI factories two years ago was hyped up, now it's real. Agent was hyped up, now there's real content happening from experts, but that tells me that it's not a strategy risk for companies with agents. It's an execution risk.
Jim Zemlin
>> That's right.
John Furrier
>> Number one conversation we went in theCUBE is, how do I execute at scale in production? That is the number one north star, pretty much 80% of the people we talk to in the real world. Your thoughts and reactions.
Jim Zemlin
>> Yeah. I think that's the right question to ask is the agent technology is real, but it has to be deployed in a serious way with all of the thoughtful consideration around security, privacy, safety, reliability, and so forth. The Agentic AI Foundation is where that conversation's happening. Why? Because all the players are here at the table, Anthropic, OpenAI, Microsoft, Google, Amazon. These are all core organizations that are contributing technology, that are adopting the protocols, standards, and technology that come out of the Agentic AI Foundation. It's where end users are coming to talk about their challenges deploying this technology. So the table is set. I think Mazin is right, time is of the essence to make sure that we bring this technology out in a way that can prevent fragmentation from happening, but I think the ingredients are here.
Mazin Gilbert
>> Yeah. And I've witnessed the AI era for decades. And I remember it took 10, 20 years for companies to be convinced AI is important to their business. It really took a long, long time. I'm not aware of any conversation we're having with any enterprise company who are not experimenting and trying to deploy agent technology. It's not taking them even 10, 20 years, not taking them even a year. It's taking them months to make these decisions, but they do want them to be secure. They have to be cost-effective. They have to scale, operational ready. These are all the right questions and that's our response.
John Furrier
>> And you know from your PhD in the days of the 80s and 90s and 2000s that AI was being worked on, but didn't have the horsepower. Now, you can have a supercomputer, DGX box for four grand at home rather than a Max Mini. That's a supercomputer basically the size of a brick. I mean, this is where we're at. That's going to spawn new software paradigms, new product innovations. We're going to have, maybe I should rename the AI factories to AI factories pumping out agents. We are going to see an agent tsunami. So what's the plan? You guys have a new conference coming out. Do you want to announce that?
Jim Zemlin
>> Yeah. So we have, in addition to the MCP Dev Summit we're doing here in New York, we're going to do that same event all over the world. We've got dates coming out in Mumbai, Tokyo, Seoul, London, in Africa. So we're going to keep the momentum going there. We also have a new AgentCon event that's coming out October 22nd in Silicon Valley. We're really looking forward to that.
John Furrier
>> AgentCon, that's going to be a flagship event for the Agentic AI Foundation?
Jim Zemlin
>> That's right. So last week I was in Amsterdam, we had 15,000 people at our KubeCon event, which is the cloud native event. This is going to be the next KubeCon. So if you're interested in deploying agentic technology in your enterprise, if you want to meet the people who are directly producing this technology and other practitioners, the AgentCon event in October in San Jose is the place to go.
John Furrier
>> Well, I'm looking forward to it. You're going to keep us busy. I have to look for a report that's going to fly around the world with all these micro-conferences. You have other news happening here at the foundation at the MCP event. Share the other happenings.
Jim Zemlin
>> Yeah. So here's another exciting thing. When you talk about agents and you talk about delegating jobs to them and having them do things with technology, one of the things that an agent has to do is pay for stuff. And in order to pay for things, you need standardized ways to do that. And so we're announcing the x402 standard is now coming to the Linux Foundation. This is technology that was created by Coinbase. And think of it as a way to make payments seamlessly over HTTP. But that's not the most important thing. When you have a standard protocol to do something, it's only as valuable as the network effect of people who are using that protocol. And so x402 has a bunch of big backers, including American Express, Visa, Stripe, Google, Microsoft, Amazon, and a host of folks who facilitate payments over the internet. So we're really bullish on x402 being announced on 4/2. So we're looking forward to seeing how we can also bring standardized ways of payments to the agentic world.
John Furrier
>> Open source continues to thunder along. Again, great example of end user participation, open sources for everyone. Congratulations on the appointment.
Mazin Gilbert
>> Thank you so much.
John Furrier
>> Executive Director, looking forward to working with you. And of course, we're launching our new agent series on theCUBE here, NYSE Wired program. Certainly New York City is the capital markets intersecting with tech. Technology is the market, the market. It's the now tech. Congratulations guys. And thank you for all the work you guys are doing. Jim, appreciate you.
Jim Zemlin
>> Thanks for having me us.
John Furrier
>> Thanks for having us on. I'm John Furrier, host of theCUBE here with big news, the new foundation focusing on agents, the Agentic AI Foundation with the Linux Foundation, continuing to advance and be a steward of open source, enabling value creation in the open with standards, doing it the right way. Thanks for watching.
Jim Zemlin, Linux Foundation & Mazin Gilbert, Agentic AI Foundation
search
John Furrier
>> Welcome back everyone to theCUBE here at the New York Stock Exchange Studios of theCUBE. Of course, we have our Palo Alto studio connecting Silicon Valley and Wall Street. I'm John Furrier, host of theCUBE. This is a preview of what's going on the Agentic Foundation. Of course, the MCP Dev Conference in New York City, the Linux Foundation is happening. A lot's going on in agents and in open source. We have two great guests from the Linux Foundation, Jim Zemlin, CUBE alumni, the CEO of the Linux Foundation and Mazin Gilbert, Executive Director, newly appointed Executive Director of the Agentic AI Foundation. Jim, great to see you. Madison, thanks for coming on theCUBE.
Mazin Gilbert
>> Thank you for having us.
Jim Zemlin
>> Thanks for having us.
John Furrier
>> Best time to be an open source. It's a whole nother level. Everyone's addicted to OpenClaw. You got open source thriving. Agents ramping up on top of Cloud Native. Amazing success with the Linux Foundation. Thanks for doing that, but a lot going on. What's the hard news?
Jim Zemlin
>> Well, so we launched the Agentic AI Foundation in December last year. And in just a few months, the organization has gotten larger than our previously largest organization, the Cloud Native Computing Foundation, home to Kubernetes, with over 170 members today, a ton of momentum. And in that three months, I have been searching for somebody to help me run the organization, and I have finally found him in Mazin here. Mazin is going to be the new executive director of the Agentic AI Foundation, and he has all the right ingredients for leading this organization. One, he has a PhD, which in the world of artificial intelligence, as you know, is a prerequisite. In fact, I think he wrote his thesis for his PhD program on neural networks. But he also has an MBA from the Wharton School. So he not only understands the tech side, which your Palo Alto office can testify to, but also this, the business side from his experience getting his MBA. But even more importantly, he spent his last five years at Google building AI solutions there. So he has all the right pedigrees to lead the fastest growing open source organization in the world. We're super, super happy to have Mazin on board.
John Furrier
>> Mazin, congratulations. I'm super excited that you're here. One, the event is happening in New York City with MCP, which is just a feeding ground for developers and innovation. But this foundation is looking from all the signals. It's going to be rapidly growing. The CNCF became really a key enabler for what we're now seeing. Took a little time to get off the runway because it was standards were growing, it was developing. A lot of hard work was done. But that work in the CNCF has set the table for the AI wave, which is highly robust, rapidly evolving at all levels. You got to pull all your tricks out of the bag on this one. Give us your thoughts and vision for the foundation.
Mazin Gilbert
>> Well, first, thank you very much for the invitation. Thank you, Jim. It's an exciting, amazing time right now. You see this space of agentic AI is moving at really the speed of light. We've never seen that kind of growth before. Companies are coming out of nowhere and coming out with some new protocols and getting adopted. Millions of people are adopting it in days or a week or two. Okay? So we've never seen this before. We're embarking into an era that I call the internet of agents. We've gone from the cloud native to the AI native era to this internet of agents where autonomous agents are going to be collaborating, reasoning. They're going to be taking actions and fulfilling those actions. We may end up with more agents in the world in the next three, five years than even human beings that we have today, which is an exciting and amazing world. The foundation is really important here, because we've seen this playbook before with CNCF. We've seen it with networking as well, with the Linux Foundation Networking. When you go from experimentation to production, you really need a neutral, open standard community that will really define the protocols, the plumbing, the safety, the auditability, the observability to make it enterprise ready. And this is what really this foundation is. And really the vision of the foundation is to create that open standard and make it real with all these participants, 170 companies and counting.
John Furrier
>> Jim and I have talked about this in the past on theCUBE and we've had many conversations at the Linux Foundation events about open source. And remember when I was in college, that was the proprietary software days and it was just open source was an underground movement at that time. So I saw that proprietary wave. A lot of us who are pioneering and open source were there and saw that growth and then everything just magically happened and here we are today. But the key inflection points at all these major ways, whether it's proprietary to open source, mini computers to open networking, standards were super important. If you look at the OSI model with TCP/IP, that was a huge thing. When you look at the internet, DNS, HTTP, certain standards that weren't restrictive, but yet open became a huge deal. I mean, cloud kind of was a standard, but I think the internet and AI kind of have a similar pattern, but at different scales, the need for that enablement.
Mazin Gilbert
>> Yeah.
John Furrier
>> Talk about that, because I think we are in a moment of consequence where at this point in time with AI and the massive innovation, that standards play, specifically standards play a key role.
Jim Zemlin
>> I think the way to think about why is agent technology so big, start there. So for decades, Silicon Valley technology companies built technology tools for you and I to use. You would log in to use a spreadsheet, you would log in to use a SaaS application. People used technology. Agents are very different than that. Agents are technology that uses other tools and technologies. That expands the total addressable market, the impact on society to a much more profound thing. And so if you think of agents as technology that uses other technology, the first question you have to ask is there are millions of different technology products and services out there. How can these agents efficiently connect to all of those different technology products and services? Well, the way you make that efficient is you have a standardized way to allow those agents to connect to that data and those technology services. And that is MCP, which is the core project and core specification at the Agentic AI Foundation. But you also need a whole bunch of other things around that core protocol. You need open source tools to allow people to create a way to deploy agents, say on their desktop. You need other adjacent technology around identity and access management, all the things that Mazin talked about. And that is what we're building at the Agentic AI Foundation.
John Furrier
>> Mazin, what's your vision on this? Because the aperture of participation, we think at theCUBE here, and I think you guys might agree, isn't it just your classic developer. There's a lot of stakeholders. AI is for everyone. It's infused in every single vertical. Vertical apps is the hottest market right now. I think that's just a predictor, a pretext, in my opinion, to massive proliferation of AI. There's a lot of even CFO conversations. Jensen Huang said token budget for your employment contract at GTC, hinting that your performance in your job will be determined by how well you use your tooling, or in this case, agents. Reminds me when sales reps had cell phone budgets, if it wasn't huge, they weren't doing their job calling customers. So you have a similar telecom AI kind of dynamic going on where tokens and for agents become not just a technical thing.
Mazin Gilbert
>> Yeah. So if you think about this, tokens to LLMs is agents are to applications. So when Jensen was talking about token budget, we're going to see a future as we get into this internet of agents is that you have an agent budget. How many agents do you have doing all sort of different type of work for you from all the kind of different verticals? If you look at the members today of the foundation, a lot of them are technology members driving the foundation protocols, the tools. As Jim basically mentioned, the training, you need that solid foundation. Once we get that and once we harmonize that, and that's what we're doing over the next weeks and months, then the verticals become really important. These are the vertical applications in finance and telecommunications and retail, et cetera. So you're going to see a lot more adoption of applications, not just number of downloads of these protocols. Like today for MCP, we have 97 million downloads a month of that protocol. Just imagine that. Okay? It's hard to believe that kind of sheer size. What we are going to go to is not just downloads of protocols. We're going to go into how many of these agents are in production and deployed. And that's really where we want to get to as a foundation, is to go from the experimentation to production to make this real,, actually solving real business problems and driving the return on investment.
John Furrier
>> It's interesting. The cloud was such a disruptive enabler and accelerator too, but the dimensions involved on the business model impact wasn't that complicated. Go to subscription model, pay by the drink, go to subscription. So easy to understand. Agents, not so much. I mean, you mentioned tokens, that's like a headcount. It's like almost a staffing challenge. It has nothing to do with technology. So less of an IT and problem that cloud solved and scaled. Now, agents are a little bit more. So it's interesting to see even some of the entrepreneurial activities are started by people that aren't even coders.
Mazin Gilbert
>> Yes.
John Furrier
>> So you have a whole nother level going here. Do you guys talk about that when you have meetings?
Jim Zemlin
>> I mean, one of the fun things about working on the Agentic AI Foundation is we get to talk to a lot of end users who are deploying agents a lot in proof of concepts initially, some in production. And we get to hear about gaps that they have as they roll that out and what they need in terms of standard methods, in terms of security, in order to make those really successful. Like for example, if you're delegating a bunch of work to an agent, how do we handle identity? Who is working on behalf of what? What identity does that agent have? You talk about cost control, right? So you have all these agents. What's the control plane in order to manage how many tokens each agent is consuming? As an agent manager, which is now a new job description, how do I make sure that I have some way of looking across all these agents and stopping one before it spends my entire token budget for the month?
John Furrier
>> Well, we had FinOps for cloud. I'm sure there'll be an AI version of that.
Jim Zemlin
>> That's right.
John Furrier
>> I mean, Mazin, this is an interesting dynamic. One of the things I was very fascinated with from a historical perspective, looking back at the CNCF and KubeCon and all the adjacency around cloud native was the participation by end users in the foundation. For example, Netflix, Lyft, Intuit. These are companies that traditionally aren't vendors selling stuff or trying to influence the market. They had real code. They were contributing some of the most consequential products in CNCF. If you take that pattern and map it to AI, you're going to have much broader end user participation because they're interested too. Open source means it's open to everybody.
Mazin Gilbert
>> And I think that's what's really exciting about this is that the barrier to entry, to contribute code, to diagnose code, to debug a code has basically lowered down significantly. Many companies who don't, like you said, don't code are now participating. And we are really looking for all these segments of companies. They're the technology companies, which traditionally are the ones who are contributing code. Then you have the large customers, smaller, medium and large customers who are already the end users who are starting to see them also contributing code. But we also care about many other companies who may not be members of the foundation or members of Linux. These are developers. These could be people in universities or colleges or in school and high school who are actually writing code or using agents to write code. And we want to grow that community as well. All of these are really an incredible sort of set of assets for this foundation.
John Furrier
>> Well, we were lucky enough to be there present at creation during pre-CNCF, watch that grow, being a participant with you guys, excited to be part of the Agentic AI Foundation. I'm sure it's going to grow very much faster. So very excited. Thanks for that. Well, we got an event happening in New York. What's the action at the MCP Dev New York activities? Give us a taste of what you're seeing and what you're announcing.
Jim Zemlin
>> We got a sold out house. So we're just off of Times Square and we have all of the core maintainers of the MCP protocol, of the Goose reference framework, of a whole bunch of different open initiatives around agents. But I think what's really interesting, and this is, I think, something important for people who are watching this to understand is the conversation around successful implementation of agentic technology is happening at a real grassroots level. And this event this week is such a great example of that. You have retail companies, Nordstrom, coming in, talking with fellow practitioners, explaining what they're seeing, explaining how they're succeeding, how they're failing. Companies like JPMorgan, you've got financial services companies here talking about agentic success and agentic struggles. And that is really what is so amazing about both this agentic movement and about these kind of events. I mean, it really, and I'm going to date myself here, it hearkens back to those kind of computing user groups that I used to go to as a kid where there was just all this incredible grassroots energy. Everybody is a builder. People are super excited. And that's what's happening here in New York this week.
John Furrier
>> It's interesting you bring that up because I was just talking yesterday on our NYSE Wired program around this new layer of persona that's developing superfast in the agentic infrastructure piece. It looks a lot like DevOps, which then became DevSecOps, which is cloud native, but it's a similar kind of culture, different people. In fact, when I wrote my post covering KubeCon EU, my opening sentence was cloud native, meet AI native. This is synergy between the two. And you talk about the numbers on the downloads, it's mass scale, but the grassroots is still an emerging community. That layer of the agentic infrastructure is propelling the AI native developer. So you have a grassroots movement at scale. I mean, when you look at that, Jim, and historically we say, "How does that happen?" I mean, usually you can't have scale in grassroots. It's either both.
Jim Zemlin
>> Yeah. I think the important thing to remember is even though we're seeing millions of downloads, rapid adoption, I think we're still in the first five minutes of a long game here. And here's why. It's easy to get quick results with agents on very specific tasks, but if you are the CEO of a company and you want to deploy agents at scales, there is a lot of things you need to think about. Security, identity and trust, access control, all of the things that you would need to really hook this up to mission-critical systems. And I think that is where the Agentic AI Foundation is creating real value, creating the standards that allow you to potentially manage a fleet of agents, making sure that security and trust is integrated into all of these openly available tools so that you can hook an agent up, not just to a low risk, maybe customer support system, but to your core financial management system. And you need a lot of different technology to do that, to make sure that private information doesn't leak, to make sure that this stuff doesn't get hacked. There's a lot of things. That is what you're going to see in the coming months coming out of the Agentic AI Foundation is a whole suite of technologies that will be contributed by the different organizations that are participating.
John Furrier
>> Mazin, any thoughts right now as you look at day one in the job? The enthusiasm's high, but the key to success in a lot of these big foundation like CNCF, what's successful was they gave a lot of confidence. You had enthusiasm and confidence in the community. Thoughts on how you see that playing out standards wise or projects? What's your thoughts on that?
Mazin Gilbert
>> We have a rare, unique opportunity and we have the best of the best at the table, the best companies in the world, the biggest companies and the smallest, and the ones that are making the headlines today, they're all the table. The funny part of the dynamics is that those are companies that competed during the day and they are really working furiously together to try to drive those standards and basically these protocols. I think time is critical here. It is the first five minutes, but the next 5 and 10 minutes matters. We have three projects right now as part of the foundation, MCP and we have Goose and AGENTS.md. The agent stack is very rich. And for us to really drive that safety, that trust, that security, there are a lot of opportunities there to bring in new capabilities, to bring in new projects, and we need to bring them fast, because projects are moving very fast right now. And if we don't do that, there will be a lot of fragmentation in the market. And yes, we will get to an exciting agent era, but it will lack standardization and it will be very expensive and that means building at scale, it will be extremely basically slow.
John Furrier
>> I mean, fragmentation encourages proprietary.
Mazin Gilbert
>> That is exactly right. And this is the best time. And I can't tell you we have two, three years. We don't. We have weeks and months. And you're going to see from the foundation over the next weeks and months, we're going to make some really exciting announcements looking at the stack from top to bottom and really coming out and bringing in some very critical projects that's going to help us sort of to really bring agents out at a scale in a secure basically way, including what are the guardrails? And when does a human come in the loop? Because agents are autonomous, but there are times you want a human in the loop. These all have to be standardized.
John Furrier
>> Yeah. I love the hype cycle, because I can tell you when the hype cycle's real because the demand for the long form CUBE content is high. AI factories two years ago was hyped up, now it's real. Agent was hyped up, now there's real content happening from experts, but that tells me that it's not a strategy risk for companies with agents. It's an execution risk.
Jim Zemlin
>> That's right.
John Furrier
>> Number one conversation we went in theCUBE is, how do I execute at scale in production? That is the number one north star, pretty much 80% of the people we talk to in the real world. Your thoughts and reactions.
Jim Zemlin
>> Yeah. I think that's the right question to ask is the agent technology is real, but it has to be deployed in a serious way with all of the thoughtful consideration around security, privacy, safety, reliability, and so forth. The Agentic AI Foundation is where that conversation's happening. Why? Because all the players are here at the table, Anthropic, OpenAI, Microsoft, Google, Amazon. These are all core organizations that are contributing technology, that are adopting the protocols, standards, and technology that come out of the Agentic AI Foundation. It's where end users are coming to talk about their challenges deploying this technology. So the table is set. I think Mazin is right, time is of the essence to make sure that we bring this technology out in a way that can prevent fragmentation from happening, but I think the ingredients are here.
Mazin Gilbert
>> Yeah. And I've witnessed the AI era for decades. And I remember it took 10, 20 years for companies to be convinced AI is important to their business. It really took a long, long time. I'm not aware of any conversation we're having with any enterprise company who are not experimenting and trying to deploy agent technology. It's not taking them even 10, 20 years, not taking them even a year. It's taking them months to make these decisions, but they do want them to be secure. They have to be cost-effective. They have to scale, operational ready. These are all the right questions and that's our response.
John Furrier
>> And you know from your PhD in the days of the 80s and 90s and 2000s that AI was being worked on, but didn't have the horsepower. Now, you can have a supercomputer, DGX box for four grand at home rather than a Max Mini. That's a supercomputer basically the size of a brick. I mean, this is where we're at. That's going to spawn new software paradigms, new product innovations. We're going to have, maybe I should rename the AI factories to AI factories pumping out agents. We are going to see an agent tsunami. So what's the plan? You guys have a new conference coming out. Do you want to announce that?
Jim Zemlin
>> Yeah. So we have, in addition to the MCP Dev Summit we're doing here in New York, we're going to do that same event all over the world. We've got dates coming out in Mumbai, Tokyo, Seoul, London, in Africa. So we're going to keep the momentum going there. We also have a new AgentCon event that's coming out October 22nd in Silicon Valley. We're really looking forward to that.
John Furrier
>> AgentCon, that's going to be a flagship event for the Agentic AI Foundation?
Jim Zemlin
>> That's right. So last week I was in Amsterdam, we had 15,000 people at our KubeCon event, which is the cloud native event. This is going to be the next KubeCon. So if you're interested in deploying agentic technology in your enterprise, if you want to meet the people who are directly producing this technology and other practitioners, the AgentCon event in October in San Jose is the place to go.
John Furrier
>> Well, I'm looking forward to it. You're going to keep us busy. I have to look for a report that's going to fly around the world with all these micro-conferences. You have other news happening here at the foundation at the MCP event. Share the other happenings.
Jim Zemlin
>> Yeah. So here's another exciting thing. When you talk about agents and you talk about delegating jobs to them and having them do things with technology, one of the things that an agent has to do is pay for stuff. And in order to pay for things, you need standardized ways to do that. And so we're announcing the x402 standard is now coming to the Linux Foundation. This is technology that was created by Coinbase. And think of it as a way to make payments seamlessly over HTTP. But that's not the most important thing. When you have a standard protocol to do something, it's only as valuable as the network effect of people who are using that protocol. And so x402 has a bunch of big backers, including American Express, Visa, Stripe, Google, Microsoft, Amazon, and a host of folks who facilitate payments over the internet. So we're really bullish on x402 being announced on 4/2. So we're looking forward to seeing how we can also bring standardized ways of payments to the agentic world.
John Furrier
>> Open source continues to thunder along. Again, great example of end user participation, open sources for everyone. Congratulations on the appointment.
Mazin Gilbert
>> Thank you so much.
John Furrier
>> Executive Director, looking forward to working with you. And of course, we're launching our new agent series on theCUBE here, NYSE Wired program. Certainly New York City is the capital markets intersecting with tech. Technology is the market, the market. It's the now tech. Congratulations guys. And thank you for all the work you guys are doing. Jim, appreciate you.
Jim Zemlin
>> Thanks for having me us.
John Furrier
>> Thanks for having us on. I'm John Furrier, host of theCUBE here with big news, the new foundation focusing on agents, the Agentic AI Foundation with the Linux Foundation, continuing to advance and be a steward of open source, enabling value creation in the open with standards, doing it the right way. Thanks for watching.
John Furrier
Mazin Gilbert