theCUBE + NYSE Wired: AI Agent Conference | Ofer Smadari, Torq

Ofer Smadari, Torq

In this interview from the Mixture of Experts series, previewing the AI Agent Conference 2026, Ofer Smadari, chief executive officer and co-founder of Torq, joins theCUBE's John Furrier to discuss how agentic AI is transforming security operations from a human-scale problem into a machine-speed discipline. Smadari explains that modern enterprises juggle 70 to 120 security tools, each generating a relentless stream of alerts — one major bank alone processes one billion security events and 200,000 alerts per day. Torq's agentic platform acts as an autonomous first responder, triaging, investigating and resolving those alerts with a 90% reduction in the analyst headcount traditionally required. The conversation also explores how AI-generated threats are accelerating the urgency for automated defense, with Smadari noting that a single individual can now build sophisticated malware in two days using AI — a task that once demanded teams of 50 and government funding. He details Torq's natural-language workflow builder, which lets security engineers describe desired automations in plain English, eliminating the need for deep coding expertise. Deployed as a cloud-native SaaS platform with onboarding measured in minutes, Torq runs billions of actions weekly for Fortune 500 customers across banking, aviation and the federal sector while sustaining triple-digit revenue growth year over year. From the evolving role of security analysts as operators rather than deep engineers to the company's "AI or die" philosophy that embeds AI across every function, Smadari provides a practical roadmap for why machine-speed defense is no longer optional.

Share this session

Clips
More from theCUBE + NYSE Wired: AI Agent Conference

Ofer Smadari

CEO & Co-Founder

Torq

play_circle_outline Promotion of Agent Conference and growing agent-builder community.

play_circle_outline Torq: SaaS Agentic SOC Automating 100% Alert Investigations, Deep Remediation, Cutting Human Responders 90%

play_circle_outline Natural Language Playbooks: Build English Workflows to Offload Engineering Tasks and Empower Security Engineers

Info
Transcript

Ofer Smadari, Torq

Ofer Smadari

CEO & Co-Founder Torq

In this interview from the Mixture of Experts series, previewing the AI Agent Conference 2026, Ofer Smadari, chief executive officer and co-founder of Torq, joins theCUBE's John Furrier to discuss how agentic AI is transforming security operations from a human-scale problem into a machine-speed discipline. Smadari explains that modern enterprises juggle 70 to 120 security tools, each generating a relentless stream of alerts — one major bank alone processes one billion security events and 200,000 alerts per day. Torq's agentic platform acts as an autonomous fi... Read more

explore Keep Exploring

What does your company (Torq) do, how did it originate, and what value does it provide? add

Can you describe Torq (background and customers) and explain how agentic systems and automation address alert fatigue in security operations — including what your product offers, how it’s deployed, typical use cases, and the customer impact? add

How do playbooks and agent-driven automation affect the engineering side of security operations (SOC)—where is the value, what should engineering teams focus on, and what is the playbook for integrating these tools into workflows? add

bolt Powered by CUBE AI

Ofer Smadari, Torq

search

John Furrier

>> Welcome back everyone, theCUBE here. The New York Stock Exchange studio of theCUBE, part of our NYSE Wired program, a CUBE original. Of course, we have our Palo Alto studio connecting Silicon Valley and Wall Street. Technology is the market. And this is our mixture of expert series and also a preview of the AI Agent Conference, agentconference.com, an exclusive event of players building technology, ushering in the agent era, Ofer Smadari is here, CEO and co-founder of Torq. Ofer, thank you for coming on remotely. I appreciate it. Hopes all is well on Tel Aviv. Thanks for coming on.

Ofer Smadari

>> Thank you so much for having me. Happy to be here.

John Furrier

>> The Agent Conference has got quite the buzz right now from Simon Chan, the community. Really, it started out as a meetup, and grew into a really active community. And the agent builders, the founders, and the technologists are leading the charge, and it's evolving super fast. So thanks for coming on. I want to expand on that agent infrastructure piece, how that's rendering itself into value. But first, talk about what you guys are doing in your company, the origination, the value, and the core products.

Ofer Smadari

>> Sure. Thank you first for having me. We are very excited to be a part of it. So Torq, we are six years old company born in Israel. Most of the employees today are in the US. We have hundreds of customers, big enterprise usually, like Uber, PepsiCo, and so on. Torq today is like an agentic security operation platform. It acts like an autonomous first responder for security operation centers. The whole job of security operation centers is to deal with the level of alerts for things, meaning that they're getting many alerts every day from different tools, and they need to defend themselves. This is what Torq is doing every day for the customers.

John Furrier

>> Talk about the agent opportunity. You mentioned a lot of alerts. I mean, I hear this also in healthcare, "So many alarms going off." It's like there's so much action happening, humans can't handle a lot of the inbound and a lot of the signaling coming in. There's a lot of signal coming in, a lot of noise as well. But I mean, as you think about agents and automation, how are security pros, who aren't new to the data game ... I mean, they've been doing this. So what is the agent piece? How does that fit in? What are you seeing on this AI-driven piece? Where are people starting? Where's the focus?

Ofer Smadari

>> Sure. So imagine that every organization is buying between 70 to 120 different security tools and infrastructure tools. Every of those are generate alerts hourly, every minute, every day. And what we are doing with agentic system is we are doing the whole investigation for 100% of the alerts from the gate. And then we are doing deep investigation, we are seeing the impact analysis, driving it to agents that are performing deterministic and undeterministic automations until resolving. The resolving can be fully autonomously, or maybe working with human analysts, and we are improving those for them to be focused on the things that you need humans to do. So if you think about it, we are building digital workers out of agentic systems today.

John Furrier

>> On the product side, what are you guys offering? Can you share the product, how the people consume it, how it's being deployed? What are some of the use cases?

Ofer Smadari

>> Sure. So we are deployed, we are a SaaS platform born in the cloud. The deployment is very easy, take minutes, usually connecting to your security tools. And then we are taking on the alerts. We are triaging them, we are building autonomous automations and agents to do the deep investigation. Then other things that sometimes need approves of humans, and then we are metering everything. We are showing dashboard of how much time took to response, how much time it took to detect those alerts. Just to give you the sense of how many alerts we are getting in, from one of our big banks that are working with us, they have one billion security events a day. Out of that, they have 200K security alerts a day. So this is the level of alert fatigue that we are dealing with. Our agentic systems are doing a human job, meaning that instead of using 200 people to do the investigation, you're doing that with our systems and a few dozens. So we are reducing by 90% at least, dealing with alerts every day.

John Furrier

>> That's huge numbers. I mean, no human can handle that. And one of the things that we expect to hear a lot at the Agent Conference in May is productivity, also workflows, remediation. There's a lot of manual process. Talk about what you guys see, and how the AIP solves those manual process issues around workflows, and then the sheer volume, velocity of inbound alerts, and the productivity. Because we've been covering security for over a decade now, and more. Burnout is huge. It's always been an issue. It's worse now, because there's a lot more happening. How is the AI piece solving that manual process piece, and the burnout, and the productivity? Can you share some data?

Ofer Smadari

>> Of course. So I believe that the only way to fight AI it's with AI, it's machine speed. So you cannot fight machines with human workflows. So I would start with the problems that AI are actually generating. I had a post about that maybe six weeks ago. Checkpoint published a post that they, for the first time in ages, one guy built a full-blown malware in two days using AI. If you asked until 12 months ago, it took an army of 50 hackers, government funding to build these kind of capabilities. So imagine that these capabilities can attack every federal organization, commercial organization like Fortune 500, and so on. So what we are building today is like a very non-static, undeterministic workflows and agents that can take almost like every alert into resolutions autonomously, completely.

John Furrier

>> A lot of people talk about productivity in terms of the SOP, the analyst, and whatnot, but there's also an engineering side of it, too. I mean, you're talking about playbooks, people are using agents to actually offload engineering. Talk about the impact to the engineering side, and where that value is, and what are people focused on. What's the playbook, I guess, for that piece of it? You got the analyst, you got the SOC, but then also the engineering side of it, because it's, workflows were involved.

Ofer Smadari

>> Of course. So the way that Torq looks, we have a broad platform from an alert into agents, workflows, and remediation. On that road, we have tools. Imagine like Cursor for security engineers. We have today tools that you just need to actually write in English what would you like to build, meaning that every time a single person is getting out of the US, and log in to his AWS, you want to create an alert, you want to block it, you just need to place it in English. And we have, today, the best agentic agents that can build these workflows and agents . You don't even need to know how to build it by yourself.

John Furrier

>> Talk about the secret sauce. I love on your site, and you guys talk about this in the market, the hyper-automation piece. Also, you have an AI or die manifesto. This is huge, because we are in it. We've crossed the threshold. I mean, AI is here. There's still issues around the identity, governance, some of the blocking and tackling, the grinding out that stuff. What is the approach? If people know that, "Hey, I want to be on the right side of this wave," what is the best practice? Could you share why the hyper-automation, and why that trend, and what are people doing to avoid the bad scenarios?

Ofer Smadari

>> So I believe that people will be more of like an operator, rather than just be very technical with deep engineering knowledge. I believe that AI will do most of the work for engineers and for analysts, and then they will be able to control it, and to plan it, and see the results, and measure everything, and plan the whole overview. The way that we build hyper-automation is super scalable. We are running billions of actions on a weekly basis for the biggest organizations in the world. They are relying on us. We are highly scaling product based on almost like every continent in the world. And my belief is that AI or die is the real thing, because in the future, I don't think that there will be a single thread of technology that will not rely on AI in some way. Most of it will be taken by AI completely. So if you are an organization and you are not adopting AI on your every day internally in your product, in your offering, I think that you can die. And we had a big SKO last week. My keynote was we are AI-first company. We are taking AI from HR, to marketing, to sale, to product innovation, always, every day.

John Furrier

>> I love the platform approach. Talk about how people engage with you guys. What's the go to market? What's the consumption side of it? How does someone get started? Take us through a day in the life. Let's just say I want to deploy this. What's the rollout? What's the engagement look like? Take us through a customer use case of onboarding. How do they get involved? Do they have to go big? Do they can start small? And since SaaS is a beautiful thing, what does the engagement workflow look like?

Ofer Smadari

>> Sure. So we have a very good sales team. Usually we are being approached by our partners. We are meeting customers in events, partners like resellers and distributors. Usually we are showing the product. The attraction is huge. We starting a POC after days or a couple of weeks. And the onboarding takes five minutes to connect, like you are signing up to Gmail. So from that on, we have a team that is helping you with your priorities to check the platform, to do the proof of value. It takes maximum two weeks until they want to continue to work in the platform in production.

John Furrier

>> Yeah. One of the things we're seeing with AI is that the freemium model's back. It never literally left, but I mean, get in quick, see it, see the magic, see what it can do, and then the adoption kicks in. Take us through some examples, what you've seen anecdotally. What's been some of the responses, from the two-week setup to deployment?

Ofer Smadari

>> It takes minutes to understand the capabilities, especially because we are replacing the first generation when we start building the hyper-automation six years ago. And now we are kind of on a third or fourth generation, because we are evolving. We are pivoting every few weeks to the next thing. And the things that I'm seeing from customers that are adopting our technology from bottom up, meaning that there are some with like a static use case, and then they are doing everything, like responding and automating identities, SOC, cloud security, you name it, everything.

John Furrier

>> Talk about some of the momentum. I want to ask you, talk about some of the highlights, customer base, revenue traction. Talk about some of the momentum you have, and put that in context.

Ofer Smadari

>> So we started in few millions first year, then we did two year, 300%. This year we are doing ... We did like 250%. This year we will do more than 200% as well. We are focusing on Fortune 500 security teams, managed partners like MDLs and MSSPs. We have fraction from, again, the myriads of the , big insurance companies, big banks, aviation companies, federal industries completely, completely like big agencies, , you name it, all over.

John Furrier

>> You got the enterprise and public sector.

Ofer Smadari

>> Of course.

John Furrier

>> All right. Well, Ofer, thank you so much for coming on, and stay safe, and I hope to see you in May at the Agent Conference. I really appreciate you, and great preview what's coming. Just for the folks watching, just put a soundbite in on agents, why they should dive into agents. What are you seeing as the value? Is there fruit on that tree if they go down the agent route?

Ofer Smadari

>> I'm with you. Agentic is the future, completely.

John Furrier

>> All right, Ofer, thank you so much for coming on. I really appreciate it. Again, stay safe. I'm John Furrier, host of TheCUBE, we're here at the Cube's NYSE studio, part of the NYSE Wired, a CUBE original program, and an open community connecting players, builders, finance, all together, coming together, capital markets and technology. Technology is the market. Thanks for watching.