Identity & access
Connect sign-ins, sessions, and permissions to the person behind the alert.

Your alerts keep coming.
Your investigations should keep up.
Turn security alerts into finished investigations. Specialist AI agents gather the evidence, test the explanations, and give your team a verdict they can verify.
Security tools keep sending alerts about unusual logins, suspicious files, and unfamiliar connections. Each one could matter, but an alert alone does not tell your team what happened or whether it needs a response.
Nova starts with the alert and gathers the evidence around it: identity history, device activity, threat intelligence, and your business context. Related facts come together in one investigation, giving your team a connected view of what happened.


Your alert queue shouldn’t decide when an investigation begins. Nova brings specialist agents together to examine identity, device, and network evidence in parallel.
Shared context. Cited evidence. A verdict your team can review.
See how Nova worksConnect identities, devices, evidence, and response decisions in one investigation.
Connect sign-ins, sessions, and permissions to the person behind the alert.
Follow endpoint activity and network signals to understand what happened.
Test competing explanations and inspect the facts supporting the verdict.
Keep proposed actions, analyst approval, and the review history connected.
Signals, context, specialists, and evidence. Follow the investigation and select a point to explore.

From the first signal to the final review, follow an investigation with its supporting evidence in view.

A backup job run by an approved administrator means something different from the same command run by an unknown account. Nova makes that organisational knowledge visible, editable, and versioned.

When an analyst disagrees, the decision becomes explicit feedback: teach a rule, suppress a pattern, or record a note. Your team can inspect what Nova has learned and spot repeated corrections.

Confidence is calculated from weighted evidence. Calibration views help your team compare confidence with historical correctness.

Response recommendations wait for review. The riskiest actions need two people; critical assets always route to a human.

Inspect the specialist, model, time, and cost of AI calls. Replay the investigation and reconcile estimated and provider-reported costs.
No. Nova investigates alerts from existing detection tools. Analysts review its evidence, challenge its conclusions, and decide how to respond. It changes the starting point of the work, while your existing tools and team remain part of the workflow.
Nova can return an Inconclusive verdict. Fixed safety checks can downgrade a conclusion when the supporting evidence is insufficient. Missing information is shown as missing, rather than replaced with a reassuring value.
Human approval is the default. The product describes five levels of autonomy set per alert type and customer, with policies that can reduce trust but cannot quietly increase it. Automatic alert closure is currently recommendation-only.
Nova is priced per investigation, with AI costs passed through at cost. Discuss investigation volume, deployment requirements, and current commercial terms in a walkthrough. No unverified price or savings estimate is presented here.
Available in the product overview a fleet view across customers, per-customer settings, and per-customer billing. A walkthrough can establish the operational and integration requirements for your customer environments.

See how evidence-backed investigations could change your team’s working day.