01

Entity Risk Dashboard

Problem

A queue of alerts doesn't tell you who to look at first. Analysts were rebuilding that priority order by hand, every shift.

Product decision

Rank entities by risk score directly, with a trend line, not just a snapshot, so a slow climb reads differently than a sudden spike.

Impact

The shift starts with a ranked answer to who first, not a queue to re-sort by hand.

02

Risk Score Breakdown

Problem

A number with no explanation isn't a tool, it's a claim. Analysts had no reason to trust it over their own judgment.

Product decision

Every score opens into its parts, alerts, asset criticality, watchlists, privileged access, each one weighted and visible. Nothing about the score is hidden.

Impact

The score became something analysts could interrogate, and therefore trust: a 27% conversion uplift in a controlled A/B test.

03

Alert Correlation

Problem

The alerts behind a risky score lived in a different screen than the score itself. Analysts closed that gap manually, every time.

Product decision

Pull the contributing alerts into the entity view directly. The evidence and the score live in the same place.

Impact

Evidence and score live in one place, removing the manual gap closing that used to start every investigation.

04

Investigation Pivot

Problem

Noticing risk and acting on it were two separate systems. Every handoff between them was a place to lose the thread.

Product decision

One click from an entity card into Timeline. No re-orientation, no re-explaining what you're looking at.

Impact

Noticing risk and acting on it became one motion: score to Timeline in a single click.

05

Risk Score Configuration

Problem

A model tuned for one environment misjudges another, and analysts had no way to correct that themselves.

Product decision

Expose the controls analysts own: engine settings, asset criticality assignments, and per-watchlist risk weighting, not settings buried behind a data science request.

Impact

Teams tune the engine to their own environment, so scores stay believable outside the demo.

06

Asset Criticality & Privileged Access

Problem

Not every anomaly is equal. The same signal on a domain admin account means something different than it does on a low-value asset, but the score didn't know that.

Product decision

Fold criticality and privileged status into the score itself, as visible, named inputs, not an afterthought.

Impact

A domain admin anomaly now outranks a lab machine anomaly, and the score reflects what the business actually protects.