Use cases

A day inside ClinivX

Six roles, six moments where the platform's guardrails and feeds do the heavy lifting. Every scenario below is a real flow you can walk through in the live demo.

Nurse

The morning med pass

  1. 1The MAR grid shows every resident due in this pass window, with overdue passes surfaced first.
  2. 2Recording a controlled substance requires picking an on-shift witness; the backend enforces a distinct second signature either way.
  3. 3A refused dose demands a documented reason; a PRN demands an indication, and the follow-up effectiveness check schedules itself.
  4. 4By 9:30 the overdue-pass feed is empty, and every administration is idempotent, so a double-tap never becomes a double-dose record.

Direct Support Professional

Documenting the shift without staying late

  1. 1After a full shift, the DSP dictates what happened instead of typing it.
  2. 2The Documentation agent drafts a structured progress note from the dictation, as a proposal, not a note.
  3. 3The DSP reviews the draft, fixes one detail, and signs. The signature is theirs; the AI never signs anything.
  4. 4Goal-linked entries update care-plan progress automatically for the next review.

Supervisor

Co-signing with actual confidence

  1. 1The pending-review queue lists every signed note awaiting co-signature, oldest first.
  2. 2Each review pins to the exact version that was signed; later addenda can't muddy what was approved.
  3. 3A note needing correction gets a reason-tagged addendum; the original stays intact, in the open.
  4. 4The audit trail records the entire chain: author, signature, review, addendum.

Compliance Officer

Staying ahead of regulatory clocks

  1. 1An overnight fall is logged; its report type sets a 72-hour regulatory deadline automatically.
  2. 2The deadlines feed shows everything approaching breach across all facilities, ranked by time left.
  3. 3The Compliance agent's morning scan files proposals for expiring credentials and unreviewed incidents.
  4. 4Closure stays blocked until every follow-up action is resolved: no quietly closed incidents.

Administrator

Next week's schedule, without the gaps

  1. 1The shift board shows next week's coverage; the Staffing agent flags a Saturday overnight gap.
  2. 2It also catches that a proposed assignee's med-tech certification expires the Thursday before.
  3. 3Overtime-risk alerts surface before the schedule locks, not on payday.
  4. 4Staff onboarding grants facility-scoped roles: new hires see exactly their building, nothing else.

Care Team Lead

Working the AI inbox

  1. 1Six agents file into one inbox: overdue passes, expiring credentials, missing documentation, care-plan reviews due, coverage gaps.
  2. 2Each proposal shows exactly what the agent saw and what it suggests: alerts to acknowledge, or governed writes to approve.
  3. 3Approving a follow-up task executes it under the approver's own permissions, through the same service a human would use.
  4. 4The run history shows every scan: which agent, when, what it produced. Evidence, not magic.

Walk these flows yourself

Every scenario above runs on synthetic data in the live demo; pick the matching role and try it.