NEXUSBY CYBERCHAIN TECHNOLOGIES
Book a Review
Menu
DEMONSTRATION ENVIRONMENTLAKEVIEW GENERAL HOSPITALSYNTHETIC DATA ONLYNOT FOR PHIFictional hospital · real Nexus governance logic

Alice AI governance

Alice · Agent profile · DEMO-AGT-004

Staff Drafting Assistant

Level 1 replacement for the discovered consumer chatbot: the sanctioned path for the same tasks, with the external cloud model allowed only for public and internal content under review.

L1 · Human with assistant

Identity and purpose

Purpose
Help staff draft internal correspondence and memos
Approved purposes
internal-drafting
Workflows
administrative-correspondence
Owner
Chief Information Officer
Status
active
Runs inside
Consumer chatbot (staff accounts) · discovered system (GOV-04 cap applies)
Registered
2026-07-01 · last authority review 2026-07-01

Evaluation on file

Status
not run
Suite
alice-eval-v1 · drafting-quality
Run
Not run
Score
Not available
Notes
No evaluation on file yet; non-public requests are routed to review until one exists.

Autonomy envelope

Human with assistant (L1): The agent drafts, summarizes, and suggests. A human performs every action that changes a record or leaves the boundary. Human role: Performs the action; reviews every output.

PERMITTED DATA
  • PUBLIC
  • INTERNAL
PERMITTED ACTIONS
  • Draft content
  • Summarize
PERMITTED TOOLS
  • none
ELIGIBLE MODELS
  • Sovereign clinical model (DEMO-MDL-SOV-01)
  • Sovereign general model (DEMO-MDL-SOV-02)
  • External cloud assistant (DEMO-MDL-EXT-01)
HUMAN APPROVAL REQUIRED FOR
  • Send outside the boundary
  • any action at risk ≥ MEDIUM
  • any irreversible action

Approver: Privacy Officer

AUDIT
  • Trust Receipt required: yes
  • Reviewer recorded on receipt: when a review occurs
  • Retention: 2 years

Decisions · synthetic day 2026-09-12

Summarize a patient chart with the external cloud modelDATA_NOT_PERMITTED · approval not_required
BLOCKOpen →
Draft a committee memo with the external cloud modelEVALUATION_UNAVAILABLE · approval pending
REVIEWOpen →

Outcomes

Governance · computed

Decisions
2 (0 allow · 1 review · 1 block)
Envelope blocks
1
Median review time
No reviews decided
Receipts
2 issued · 1 with evidence

Business value · synthetic measurement

Consumer-chatbot sessions on clinical subnets
0 vs ~40 / week · Last 3 weeksProxy logs after the block and the sanctioned path

← All agents

Alice on this site runs over the synthetic Lakeview General Hospital environment. Verdicts, checks, and receipts are computed by the Nexus decision engine on synthetic requests; nothing here is a customer, a patient, or a production deployment. Signed-in organizations see their own agents, decisions, and receipts.