Roy Zisapel
Analyst · Jefferies.
Yes. So I think in our prepared remarks this quarter and last quarter, I mentioned some areas that we're using AI as part of our complete solution, what we call the EPIC AI framework, and we're infusing AI to each and every layer of our solution. So already from the protection engines themselves, meaning the algorithms we use to detect and mitigate, I was talking about how we are detecting and mitigating business logic attacks. We're leveraging their AI. Our competition over here is very manual. You need to define a lot of rules, your business logic, it changes as application change. While we really built in real time an LLM dedicated to your application, dedicated to your traffic, it provides significantly different results in configuration, in onboarding, in detection, in mitigation. That's one example of how do we use AI in that protection engine to differentiate. Then I mentioned in today's comments, another layer, which is the SOC-X AI. We've built on top of our platforms, AI agents. And as far as I know, the competition does not have it today, especially not in GA. And those AI agents are sitting on top of the platform and watching your traffic. If they detect a suspicion that an attack has passed the controls, we would automatically, the AI agent would do an analysis of your traffic, would come to a conclusion whether it's really the case that attack is leaked or maybe a flash crowd coming to your website. And if it's an attack, we will automatically come with your recommendations, what to change in order to mitigate the attack. In our competition, all these -- and also until, I think, 6 months ago for other, all these activities were done by SOC people, by humans and only collecting the information, analyzing that, getting into recommendations, testing them. This is a long cycle. It can take from, I don't know, half an hour or 2 days. We are shrinking that time frame to seconds and minutes, delivering significantly better mean time to resolve for our customers, which means less downtime, less risk, less exposure, significantly better business outcomes. So this is, again, a big differentiator for us. And as I've mentioned, it won us several customers, both displacements and new. And last but not least, there is the whole area of protecting AI applications, which we started to discuss in this call, which is another area. That's early. I cannot talk yet about competitive advantages. Let's see in the coming months who comes with what solution, but we feel very strong based on our heritage in algorithms, based on our experience in AI security for years that we will be very strong there as well. So that's in a nutshell, how we use AI to differentiate in the market of application security and dealers.