How do you or your organization use this solution?
Please share with us so that your peers can learn from your experiences.
We use Vectra AI to sniff the network using Ixia taps so that we can identify potentially malicious activity on the network and at all points of the kill chain. What it's really good at is correlating seemingly unrelated events. It's in our data center, but the versioning is controlled by Vectra. They push it out discreetly so I don't have any touch on that.
We have two use cases. The first is that Vectra's platform allows us to get visibility into anomalous behavior, which, previously, we never really had access to, for threat hunting and incident response. We use it in support of our incident response operations to help supplement our investigations on hosts. We use it to correlate any suspicious activities, which is something that Vectra has been extremely accurate in, when used the right way. The second use case is that we've used the Vectra Cognito Recall and Cognito Stream devices. With these integrations, it's given us instant visibility into all the network data as well. That enables us to conduct our own hunts on our network data, data you'd see on a SIEM solution. It also gives us the ability to correlate with our playbooks because it gives us access to the data itself in much more depth and detail.
Our main intention was to see what type of visibility, in terms of detections, Vectra could give us. We use it on both our manufacturing perimeter and at the internet perimeter. That's where we have placed the devices. We have placed it across four sites, two in UAE and two outside UAE.
We use Vectra with the assumption that our other defensive controls are not working. We rely on it to be able to detect anomalous activities on our network and trigger investigation activities. It's a line of detection assuming that a breach occurred or has been successful in some way. That's our primary use case. We have it in some of other use cases, like anomalous network activity and detection for things. E.g., we are trying to refine or improve suspicious internal behaviours because we are a development technology company. We have developers doing suspicious things all the time. Therefore, we use it to help us identify when they are not behaving correctly and improve our best practices. We have it predominantly on-prem, which is a combination of physical and virtual sensors. We also have a very minor element on the cloud where we are trialing a couple of components that are not fully deployed. For the cloud deployment, we are using Azure. We are on the latest version of Cognito.
The original use case was because we had some legacy stuff that doesn't do encryption at rest. Compliancy-wise, we had to put in some additional mitigating actions to protect it. That was the start of it. Then, we extended it to check other devices/servers within our network as well. We are on the latest version.
One of the biggest things is the visibility of stopping or identifying any infection as soon as possible. In this case, if someone downloads something malicious to their workstation, we have a number of controls in place. However, it wasn't so much the endpoint. It was the spreading of a worm type scenario or a WannaCry type thing. Anything that could potentially spread after the initial infection, which is where we wanted to come in and get that visibility. It was key for us to have something that we could use for identifying as soon as possible, which would be call center initiated. That was probably our biggest thing: To push it in that direction, as we're a regulated company from the FCA. They drive us continually for improvement and behavioral analysis. Network analysis sort of falls into that bucket. We already have a SIEM, which some people would argue gives us a lot of that visibility. It doesn't tend to give it the focus that we need. From Vectra, we get a lot of alerts of, "This is happening," or, "This is unusual." This is a lot easier than waiting for a couple of logs to come in, then a bit of AI logic at the back of it to potentially push it in that direction. It's very much for us to get a view of a potential attack, then deal with it as quickly as possible. To pinpoint where it's coming from, and where it is going to go. One of the biggest things that I wanted to ensure is that it covered our call centers because that is where I see my biggest risk. So, I was really key on getting sensors across all geographic locations within the UK and in all of our small communication rooms. It is all on-premise. We have a number of call centers spread around the UK. We look at all east-west traffic, as well as north-south. It all goes into our brain in our data center. We do have some branches out in Azure, but we're waiting on the new plugin that they are trying to develop. We are just starting in on our cloud journey and most of our infrastructure is in still private cloud. We haven't really gotten to the point where we have public cloud. We're up-to-date, but I don't know the exact version number that we are on.