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
Noam Schwartz, Alice
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.
In this interview from theCUBE + NYSE Wired: Mixture of Experts series, Noam Schwartz, chief executive officer and founder of Alice, joins theCUBE and NYSE Wired's Gemma Allen to discuss the company's evolution from a content-moderation powerhouse into a full-spectrum AI safety and security platform. Schwartz explains how Alice, formerly Act Defense, spent years protecting major platforms against fraud, impersonation and malware before a pivotal shift in late 2022. As foundation model labs began seeking help with model hardening — ensuring AI systems cannot b...Read more
exploreKeep Exploring
How did the company evolve from protecting large user‑generated content platforms against long‑tail harms to offering AI model‑hardening services (culminating in the Alice product)?add
What are the main types of customers the company works with, and how does it help them harden their models for accuracy, safety, security, privacy, and governance?add
How are threats coming through AI systems, and what are the primary entry points (gates) for those threats?add
How do you build and maintain customized content-moderation and AI-safety models for clients?add
What should we expect from you and Alice over the next year following the rebrand—your plans, priorities, and any major milestones or product developments?add
>> Welcome back to theCUBE, coming from our studio here at the New York Stock Exchange. This is our mixture of expert series, one of our NYSE Wired programs, where we're connecting Silicon Valley to Wall Street. And joining me now is Noam Schwartz, CEO and founder of Alice. Welcome, Noam.
Noam Schwartz
>> Hello. Thank you very much for having me.
Gemma Allen
>> Thanks for being here. So Alice, newly named. I have always known the company as Act Defense. Maybe start there. Talk to us about the recent rebrand.
Noam Schwartz
>> Okay. So that was long in the making and it's supposed to represent a massive change that we had in the last few years that finally came into the final conclusion with Alice. So what we did since 2018 when we were founded is protecting big tech, large user-generated content platforms against long tail of harms, impersonation, fraud, malware, like the worst of the worst. And the way we did that, by collecting data from really bad sources of chatter, analyzing that, sourcing it, merging it into models that can detect content at scale. And we are very good at that. We're the best in the world. Now, in 2022, something big happened in the world. AI moved from the laboratories to the real products. It's now mainstream. And around that time, Q4 2022, '23, we got a lot of market pool to help the AI foundation labs do something called model hardening, which in very simple terms means help us control the AI so it won't be abused, so nobody could do bad stuff with it. Nobody can hack it and nobody can use it to cause harm to other people or other organizations or to the company that is creating it. And this thing started as kind of like a science project almost, but then it became like multimillion-dollar business unit and it became multi-tens of millions of dollars of business unit. And we thought, wait a second. It's not a different concept. It's like a different story. Something is up. And we realized that because of AI, not only that it's easier to cause harm to the big platforms and it's cheaper, it's faster, it's more accurate-
Gemma Allen
>> It's more user-friendly, right? In some respects.
Noam Schwartz
>> Yeah. Everybody can be a bad actor right now, which is a massive danger for our regular customers. But also, AI became the threat itself. It became a new door that you can get into an organization that is using it and actually cause harm to that organization. And in the future, and we're already seeing the green shoots of it, AI can turn against the company that is using it, scheming, overreaching and these kind of things that sounds futuristic, but they're actually happening. So we thought there's a new story here. We need to think about who we are, what do we tell our customers, what do we tell the market? What do we tell ourselves? We're now adopting something that is completely new, a completely novel technology. And we thought that AI is something so unique that it can be terrific, create incredible benefits, but it also can be terrible. It can take the world to like very weird places. There's like this door, this drink, Alice in Wonderland came to mind, and this world of connotations really clicked with us. So we became Alice.
Gemma Allen
>> I didn't realize-
Noam Schwartz
>> Formerly known as Act Defense. The brand of Act Defense, our work with a lot of the big platforms remain as a very important piece of the puzzle. It's the same team, same values, but Alice is our future.
Gemma Allen
>> Wow. Well, we're certainly not in Kansas anymore anyway. That is for sure.
Noam Schwartz
>> It's a different story, but-
Gemma Allen
>> So talk to me a little bit about the product in terms of the problems that you're solving, right? Because absolutely, this is no longer just about people being able to impersonate from the perspective of fraud. It's also about a massive cultural shift and how technology is being weaponized across all aspects of society. And one thing I always follow on your content, Noam, is you talk a lot about the cultural aspects as a dad, as a member of society, in terms of like how we raise children on these platforms. There is a lot of people, including myself, who are really concerned about what this future would look like. And would you say that this is content mod as a service in the AI realm, or is it something more complex than that?
Noam Schwartz
>> Yeah, it's something fundamentally different. So first, in terms of like how society is changing right now, we need to separate between how our kids are growing up, what kind of content are they consuming, how even democracies are going to change because of AI, and how the workplace is going to change, because those are kind of like the two different, let's call it narratives out there. In terms of society, well, this is not a new thing. We've been through social media. We've been through user-generated content. Now we're seeing AI influencing the habits of our kids, the habits of content consumption. I don't think this would be as fundamental as people think. There's going to be more content, more choices, but effectively, I don't see something really important. In terms of how the workforce would change, think about that when folks moved from using type machines to using computers, it wasn't easy to a lot of people. And now everybody's using the state of the art, Salesforce and HubSpot and Google Suite and like these are kind of like the tools of the white collar worker. They also stopped along the way in Windows 3.1, Windows 95. And how happy the people that mastered that are now that they need to master way more complicated tools. So people that would actually make this shift, this change will win in the next few years. And I'm also telling that to my team. You need to follow the technology, you need to figure out how to actually use those tools, and you are saving your careers. It's not about you're going to be replaced. You must learn how to use these tools to be relevant.
Gemma Allen
>> My mother was infuriated by the invention of mobile phones, right? She just saw such a useless product. I look back at that. I'm like, that is so hysterical retrospectively, right? But you're right.
Noam Schwartz
>> Yeah, I get it.
Gemma Allen
>> Cultural adjustments need to happen in a technological change. So talk to me about the workflow. So I read that you guys are operating or working alongside seven of the largest frontier models globally.
Noam Schwartz
>> Yeah.
Gemma Allen
>> That's impressive. So in some respects, it's kind of poacher turned gamekeeper from the perspective of what's happening in the space, right? Because some folks claim that there isn't enough moderating happening at a platform level. I mean, Anthropic have had a very interesting week, and I know it's not necessarily about moderation, more about the use of the technology, but it's along the same lines of commercialization versus regulation and ethical intent, right? So what sorts of examples or workflows or use cases are you seeing when you're working along these big players?
Noam Schwartz
>> So we work with two main types of customers. We work with the big players, the creators of the models, the creators of the, let's say, the producers of the tokens. Those who really, really care that it will be accurate and it will be safe. They care about the safety of those things, safety and security and privacy and governance because they want these tokens to be adopted by the enterprise. They're creating it for a reason. So we help them make sure that their models are hardened, which means, again, in the simplest terms possible, models are continuously being fine-tuned for accuracy, to do the work they were designed to do. And they're also constantly being fine-tuned for safety, to make sure they're doing just the work they were designed to do and nothing else, that you can't jailbreak them, that they are not going to produce harmful code, novel malware, they're not going to be abused for cybersecurity incident. They're not going to help you impersonate someone else. So we do a lot of work with the big companies to help them do that and to make sure that everything they produce, if it's a model for chat or if it's a model to create agents, it will do exactly what it's supposed to do and nothing else. Think about, we're the company that helps you protect from the unknown unknowns almost of AI. On the other side, in the enterprise, enterprises are now adopting AI like nothing else. The introduction of Claude Code changed the world way more aggressively than everybody intended.
Gemma Allen
>> I mean, look at what happened this week on the stock market, right? Look at some of the reaction to some of the Claude Code announcements.
Noam Schwartz
>> It keeps happening.
Gemma Allen
>> It's insane.
Noam Schwartz
>> It's controlling the narrative, and it's beyond developers. It's like, people can't imagine how code beyond intelligence is no longer a limiting factor of progress. So if everybody can code, everybody can develop, and software is something that we're producing like documents, what does that mean for safety and security? What does that mean for guardrails? If we're producing so much code, how do we make sure that it's the right code, that it's aligned with our interest, with our KPIs, but also with the security and safety requirements of the organization? And this is where we fit in. We help enterprises deploy runtime guardrails, automatic testing that is based on what they're trying to achieve. And when you're combining that with the work we do with the foundation models, you just got the most powerful safety and security AI engine in the world.
Gemma Allen
>> Let's talk about the speed of that interception, right? On the technical workflow around it. Are you saying... I'm familiar with the content mod version from Meta like maybe six years ago, right? Where you had a number of people, there was sign-off layers, there was like escalation points. Are you in a place whereby this technology is training and fixing itself? Is it intercepting live events and shutting down accounts or shutting down instances? Or how... Tell me about the volume.
Noam Schwartz
>> So exactly like that, but instead of user-generated content, AI-generated content. So we created a new category called conservative conversational tech, which basically mean that, think you're going to a telemedicine company and you're asking questions like, "Hey, I have this and that symptom. What is it?" And someone needs to scan that content to make sure that it's based on the company's policy, that you're not trying to hack into someone else's database or you're not trying to cause that model to do something it's not supposed to do. And then when it does its thing, analyzing your text, operating all kinds of tools, accessing databases, and then providing you with the answer, you want our technology to scan that answer and make sure that it's based on the policy, it's accurate, it's not containing anybody else's data. So it's like moderation, but nobody, no human is viewing it. It's only based on other models. So it's like, as you said, it's like self-supporting mod, but in a way bigger scale with more topics and much faster, like 200 milliseconds.
Gemma Allen
>> Wow.
Noam Schwartz
>> Yeah.
Gemma Allen
>> And tell me about the sorts of vulnerabilities or threats you're seeing, the landscape for those, right? Is it coming, especially in enterprise, I'm very interested in this, is it coming from internal? Is it staff? Is it folks that are perhaps searching or parsing sites that they don't realize the impacts? What sorts of ways are these threats and these falsities touching enterprise? How are they coming through? What's the gate?
Noam Schwartz
>> So there are like three main cases. There's the adversarial, there's the naive and there are the mistakes. The adversarial use cases is there's like a threat actor or just like someone who's trying to like go against the company and they're trying to do like new types of phishing and trying to manipulate the model to get data that it's not supposed to give you. So it will send a PDF with malicious instructions. They know that there's piece of AI that is scanning that PDF. Let's say that you're sending like a job application for like a big corporation and there's no longer a recruiter that scans everything manually. There's a model that scans everything automatically and then they prioritize. They try to manipulate that model either to get them to the top of the list or just to understand who else submitted the candidacy or just to pull some API key that they're not supposed to pull. So those things are malicious. Bad actors are trying to do that. Another way is like a mistake. An employee just played with some chatbot, some agent that was created in the organization. That agent went online, went to a website with untrusted information. That agent can also call all kinds of internal tools and delete data, and lo and behold, a simple mistake cause the organization to lose a lot of data or even to send data it's not supposed to send to like a third party. And there's things that are happening because of pure mistakes, naivety. Someone created a new application using Claude Code or any other of the tools that can create code. And that was not a very elegant code. It worked. It wasn't elegant. And after two weeks, they discovered there's a vulnerability, a very known vulnerability in that code, and you can just go ahead and access the server and take whatever you want. So all of those things happening all the time. And there's so much code right now. There's so much AI that CISOs don't know what to do.
Gemma Allen
>> Talk to me about your own stack. Is this completely proprietary? Do you use many of the models? Talk about the build.
Noam Schwartz
>> So we don't use any of the models. We use our own models. So in the past like eight years, we built a lot of small models that are able to detect content at scale. So they are like cheaper, they're faster. They're tuned in on the specific violations and the policies that we're looking for. So they're not like general. We also have the ability to produce models that are fine-tuned to the policy of the organization that we work with, which is a very, very unique capability. We work with like, let's say a new company, a pharmaceutical, a bank, a gaming company, an entertainment company, and they basically define, this is what good looks like, this is what bad looks like according to us. And then we harden that with our models and they get their own strain, their own DNA of model that is exactly what they need without the false positive that we're seeing with like other guardrails in the market and other solutions, and then they can actually trust their model. Now, on top of that, we have the largest security reverse engineering AI laboratory outside of big tech in the world. So think about 100 people constantly looking for new vulnerabilities in new AI models and they are working day in, day out, finding the new things, understanding the new research and implementing everything into our models. So those things working like hand in hand and creating this synergetic value that provides our customers the best value they can get.
Gemma Allen
>> Talk to me about the speed of keeping a pace with this market and these dynamics, right? So that's impressive to hear that, but I would imagine that companies aren't working on 12-month sprints anymore. You're working on probably like almost daily sprints in terms of the constant updates. It seems to me as though there's a launch or a re-release almost every week. From an R&D perspective, what does that mean, like when you're running a company that's kind of trying to keep a pace and use it as a, I guess, defensibility sell back into these markets?
Noam Schwartz
>> So there's a lot of uncertainty in the market today in terms of what the market really needs. We don't build a year in advance. Our product roadmaps are... Nobody thinks that we can estimate what will happen a year from now. It's like a few weeks or months, but also the development velocity is shorter, so we can actually do that. Right now, a lot of enterprises are looking at themselves and asking, are we moving fast enough? Are we using the right technology? Because one day, OpenAI release something and then Anthropic releases something and then Google is releasing something and then Grok is coming from the left side. And like, what's better? What's most secure? It's very hard to know what's the right move. And the good news, the entire industry shares that. But the way that we operate is small teams, shoulder sprints, listening to our customers, but also be able to define what good looks like and not wait for every single to arrive, because technology is moving way too fast. And the adoption level is way slower than the introduction of new technology. So as you said, every day, there's something new, and I don't expect the enterprise to adopt that fast enough. And our job is to help them adopt faster, because every CEO, every board, they want to see the organization increasing productivity and doing more stuff and moving faster, But they're very afraid and also they don't know how to do it. And that's where we fit in.
Gemma Allen
>> Well, every day is full of new surprises now. Tell me, close us out. What's ahead for you and Alice, this rebrand? And I know we can't talk about a product roadmap from a year out, but tell me what's ahead for the next year for you and the team.
Noam Schwartz
>> So we're going to cross $100 million of ALL, which is like a big thing for us this year.
Gemma Allen
>> Congrats.
Noam Schwartz
>> We're working with terrific customers. We're introducing novel ways to defend AI. We're looking forward to Agentic AI to really take over the entire industry. There's incredible developments that are coming from the foundation models that were not introduced yet, and they're going to shift the way people even understand the potential of the technology. From OpenClaw that opened the mind of so many builders around the world and created the viral phenomena that is just getting started, and that's January. We're going to see a lot of development this quarter. We're going to introduce a very interesting new product that is going to change the industry in Q2, and sky's the limit.
Gemma Allen
>> You guys going to be at RSA?
Noam Schwartz
>> Yeah.
Gemma Allen
>> Great. Okay. Well, so is theCUBE, so maybe see you there.
Noam Schwartz
>> You won't be able to miss us. It's going to be a massive rabbit, a massive clock, and we're going to have a mad tea party.
Gemma Allen
>> Good. I love a good tea party. No, thanks so much for coming on theCUBE, NYSE Wired.
Noam Schwartz
>> Thank you very much.
Gemma Allen
>> I'm Gemma Allen, coming to you from the CUBE Studio here at the New York Stock Exchange. This is our mixture of expert series with NYSE Wired. Thanks so much for watching.
Gemma Allen