businessinfomag.uk magazine 32 automation, your management, all those challenges, so you end up with other platforms. We try to integrate as much as possible, but over time a lot of our vendors and MSPs have said how can you make this easier for us. Cyber Frame was a natural development for us because we have built infrastructure for 40 years – our disaster recovery is purely infrastructure to replicate what you have somewhere else. We have 50 data centres around the world where you can deploy Cyber Frame, and the moment you deploy it you don’t have to worry about protection, because it’s built‑in and natively integrated. That’s simplicity. You can also deploy it on-premises, with the same infrastructure management, the same agent. The whole sovereign movement, where people say ‘I have infrastructure in the cloud, but I would like to have secondary infrastructure on-premises’ is the reason we started Cyber Frame. There’s another reason why this is very important, which is that the big five control the vast majority of AI infrastructure. In April, we launched GenAI Protection for the simple reason that people have no knowledge, no control, no management capabilities and no idea what their employees are doing in AI and no idea if private information is going to be used for training. That all comes back to not controlling the infrastructure. GenAI Protection puts a layer on top that basically gives you visibility into employees’ use of AI – what tools they are using, how much they are using them, whether they have the right settings, whether they are sharing confidential information. It can also block confidential information from being sent. In the future, people will want to own their own infrastructure running an LLM model with local control. With Cyber Frame, we can start serving GPU-type infrastructure on-premises, letting customers run their own models for training using their own data set within their company. It’s at a very early stage but I think the potential market is huge. www.acronis.com demanded. Twelve months ago, we built a three-year roadmap and half of that roadmap will be delivered six to 12 months earlier than we had planned. The biggest opportunity for our MSPs is how they drive their business, because MSPs are some of the most challenged service providers out there. Normally, a service has 30% margin; MSPs take between 10% and 20% margin, so they’re working on the edge. Alt-Tech manages 500 endpoints per technician. That’s already double the industry average of 200 to 300 endpoints and in the next 12 to 18 months it expects to reach 1,000 endpoints per technician. Their efficiency will help them win more customers and serve customers better because they will have more time to spend with them and build trusted relationships. JG: How is Acronis using AI to enhance its own operations? JJ: We are doing a couple of things. One is in finance. Normally we close out the month and then do all our calculations, reporting and so on. That can take anywhere between six and 10 days. We now do that in three. The outcome is still the same; we just do it faster. Engineering is probably the best example. When we started using AI for engineering, we saw that people very quickly became more productive, generating not 100 lines of code, let’s say, but 150. Projects that used to take six months can now be done in four, allowing us to launch features and functions faster. This is why it is so important not to use AI to reduce your workforce but to enable you to do more in less time. Our main goal is to use AI to develop new solutions faster and to build the protection customers need more quickly, because adversaries will also be exploiting AI to go faster. JG: How does Cyber Frame fit into your AI-first approach? JJ: Today, everything goes in the cloud and we protect it. You build your cloud infrastructure or you go to a cloud provider and then you connect your disaster recovery, your back up, your security, your making another dashboard or a report to help you reach a decision. AI should be making that decision for you. If you have AI connected to your core platform and to all your different services, it can help you to become more autonomous. That’s what we call Cyber Intelligence. It’s not about the information or the intelligence that you get out of it, it’s the first step to becoming autonomous, because with intelligence you can create an action. If I see that a particular customer has been subject to more spam or more attempted hacks I would normally go to the client and say this is what our intelligence shows, I suggest enhancing your email security or endpoint security to prevent an attack from succeeding. With Cyber Intelligence, you don’t have to do that. It will say this is the action I would take based on the information I have. Would you like to execute? And you can click on it, and it executes automatically to make sure that the user is protected straightaway. AI Studio lets you define how much control you want. For any step of the process, you can add a manual intervention to check and give the OK to move on, but that will not always be necessary. For example, a recurring problem in the MSP world is poor endpoint performance, usually because a task has taken a lot of resources and just needs to be killed, rebooted or changed. Often a nontechnical person will call the MSP who will then look into it. With Cyber Intelligence and Cyber Studio, you can build a workflow with an agent that basically says we’ve seen this before and starts executing automatically without anybody being involved. Action rather than just information. JG: Is it your vision that ultimately a machine will be able to do everything autonomously? JJ: I don’t think machines can do everything. There are still certain things in life today that can’t be done by an agent. When we optimise our processes, we’re not trying to reduce costs but to do things faster, because that’s what’s Q&A …continued
RkJQdWJsaXNoZXIy NDUxNDM=