Owning a detection product and running a detection function are separate purchases, and only the second one is any use at two in the morning. Every line here ships with the agent on the machine, the correlation sitting above it, and the people who decide what gets done. Nothing in it is a screen that somebody quietly hopes you will watch.
Recognising a known sample has not been enough for years, and it was never worth much on hardware that spends its week nowhere near your network. The SentinelOne agent builds a picture of what a program is doing where it is doing it: what it launched, what it opened, who it called, and whether the resulting shape reads as scrambling files, hoovering them up, or somebody moving sideways with no hurry at all. None of that reasoning needs a connection, exactly the property you need on a laptop over the Atlantic and on a terminal wired to a packing line.
Fluency then ties what the agents saw to everything else in the building: authentication records, message events, traffic, and logs from products you are already paying for. A case that arrives at our desk carrying its surroundings can be decided in minutes. A case that arrives on its own becomes a research task, and research tasks form a queue.
The first line sorts and recommends. The second opens up what gets pulled together, so a peculiar authentication on one continent and a strange binary on a laptop somewhere else cease to be two unrelated curiosities logged by two unrelated people. The third acts without being asked, which suits hardware where an hour of waiting costs real money.
Containers are metered separately for a reason. A node is not a desk, the agent behaves differently on it, and counting a node as though it were a desk would produce an invoice that lies wrong. Meter nodes. Never meter pods.
Every rate below is fetched from billing the moment this page opens. Whatever gets marked waits in the schedule; reading on costs you nothing.
Conviction happens on the machine. Judgement happens at a crewed operations desk. You are sold both. That is why a finding turns up recommending something, instead of as a chart somebody expects you to decode before your first coffee.
| Platform | SentinelOne, correlated by Fluency |
|---|---|
| Applies to | Windows, macOS, Linux |
| Response | Told, advised, and helped to remediate |
| Without a link | Judgement continues with no network present |
| Desk | Crewed at Fortify 24x7 every hour of the week |
| Measured in | Protected endpoint, each month |
Take the previous line and open the aperture. Sign in records, message events and network flow get read beside what the agent saw, rather than sitting in three consoles owned by three people who compare notes on Thursdays.
| Platform | SentinelOne, widened through Fluency |
|---|---|
| Draws on | Machines, identity, messaging, traffic |
| Response | Told, advised, and helped to remediate |
| Archive | Deepened, so a window can be re-examined later |
| Suits | Estates already gathered into one Microsoft or Google tenant |
| Measured in | Protected endpoint, each month |
The widened tier, given permission to act. Cross the threshold and the machine is cut off and put back the way it was while an engineer is still opening the case file, not once the case file has been read.
| Platform | SentinelOne, acting unattended |
|---|---|
| Cutting off | The affected machine loses its network interface |
| Putting back | Changes made by a ruled against process are reversed where supported |
| Audit | Each unattended action is checked by a person afterwards |
| Suits | Finance workstations, plant controllers, signing hosts |
| Measured in | Protected endpoint, each month |
The same discipline pointed at containers, and metered by node so the bill agrees with a figure your platform team can quote off the top of their heads.
| Platform | SentinelOne for Kubernetes |
|---|---|
| Watches | The behaviour of workloads once the scheduler has placed them |
| Response | Told, advised, and helped to remediate |
| Desk | Crewed at Fortify 24x7 every hour of the week |
| Measured in | Kubernetes node, each month |
Node coverage, aperture open. Cluster activity gets read alongside authentication and machine evidence, so whoever is on call is not investigating in a vacuum at three in the morning.
| Platform | SentinelOne for Kubernetes, tied together by Fluency |
|---|---|
| Draws on | Node runtime, identity, machines, traffic |
| Response | Told, advised, and helped to remediate |
| Archive | Deepened, so a window can be re-examined later |
| Measured in | Kubernetes node, each month |
Node coverage cleared to act, for clusters where something misbehaving overnight is not survivable until the first laptop opens on Monday.
| Platform | SentinelOne for Kubernetes, acting unattended |
|---|---|
| Shutting down | Unattended action against whichever workload misbehaved |
| Audit | Each unattended action is checked by a person afterwards |
| Suits | Production clusters carrying work your customers can see |
| Measured in | Kubernetes node, each month |
Written down in advance, so the question of what else the estate needs gets asked now and not in the middle of something.
Heads up: card statements show FORTIFY 24X7 - Iconic IT Innovations is a Fortify 24x7 brand, and your subscription is billed by Fortify 24x7.