Workplace investigations that intake, route, and wall themselves off.

The configuration that makes investigations run safely: the allegation taxonomy, accommodation and evidence categories, confidential intake paths, case routing, the access controls that keep sensitive cases need-to-know, performance-improvement lifecycles, and case reporting.

The cases, allegations and interviews themselves stay with your team — Phyllis configures the engine that intakes, routes and protects them; she doesn't manufacture the records your agents work.

Confidentiality is the whole point of ER, so she treats it as a first-class part of the build. She reads who can currently see cases and stages the access controls that keep sensitive matters restricted to the right people.

She also watches for the classic leak — notifications that quietly carry sensitive detail to people who shouldn't have it — and closes those gaps as part of the design.

Yes, and she recommends the right built-in mechanism rather than a fragile custom form. For true self-service anonymity she proposes a reporter-facing intake where identity is never recorded and the case is tracked by a secret key. For phone or walk-in reports she uses the built-in agent-assisted path that keeps the reporter anonymous.

Correct. Phyllis explores, proposes a plan, and only builds on your explicit approval, with every change staged for review. Given how much ER touches confidentiality, she keeps every access change visible and staged — nothing is applied until you approve it.

Every claim is cited to where it came from in your own instance — the allegation categories already defined, the groups that can currently open a case, the intake feeding your queue. She reads your live setup rather than assuming out-of-box defaults, because on migrated instances the taxonomy and access rules drift from the shipped baseline.