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

Alice AI governance

Alice · the authority layer inside Nexus

Alice decides what AI is allowed to do.

Nexus sees the environment. Alice governs the decision. Trust proves what happened. Alice is the authority layer inside Nexus: a policy decision point that evaluates every governed request against identity, purpose, data, model, tools, policy, risk, and the agent's autonomy envelope — and answers ALLOW, REVIEW, or BLOCK with the evidence attached.

12Decisions · Sep 12
3Allowed
6Held for review
3Blocked

One loop, three layers

Nexus sees

Intelligence and readiness

What is changing outside, and where the organization is exposed inside.

Alice governs

The decision

Every governed request is evaluated the same way and answered with structured evidence.

Trust proves

The receipt and the outcome

What happened, why it was allowed, reviewed, or blocked, and what it produced.

What Alice evaluates

Thirteen inputs, one structured verdict. No hidden reasoning: every check is shown with its status and reason code.

Human identityAgent identityApproved purposeBusiness workflowData classificationModel eligibilityTool permissionsRequested actionPolicyRiskAutonomy envelopeEval resultsRuntime context
ALLOW

Every check passed inside the agent's autonomy envelope. The action may proceed and a Trust Receipt is issued.

REVIEW

The request is held for a named human. Nothing executes until the reviewer decides; the hold itself is recorded.

BLOCK

Refused by policy or envelope. An eligible alternative is offered when one exists. The refusal is evidence that the control worked.

ALICE IS A POLICY DECISION POINTAlice decides; enforcement points enforce — gateways, identity providers, data-loss-prevention tools, proxies. Each one is labelled honestly on the integrations page: 2 live signal sources, 4 built and demonstrated components, and the enforcement points that are planned or conceptual.

A decision, end to end

Summarize a patient chart with the external cloud model: Hospitalist, Internal Medicine via Staff Drafting Assistant.

  1. 1
    RequestPHI · model DEMO-MDL-EXT-01 · network external
  2. 2
    Alice: BLOCKDATA_NOT_PERMITTED — PHI is outside the envelope (PUBLIC, INTERNAL).
  3. 3
    AlternativeSovereign clinical model is eligible for this request inside the envelope.
  4. 4
    ReceiptSealed as chain event; chain verified: yes. Open the decision →

Agents under authority

Discharge Summary Assistant

L1 · Human with assistant

Draft discharge summaries from the chart for clinician review

pilot · owner Chief Medical Information Officer

Scheduling Optimizer

L2 · Human-agent team

Propose and apply shift patterns from historical staffing demand

active · owner Director of Operations

Nexus Intelligence Analyst

L3 · Human-led, agent-operated

Collect and summarize public healthcare-AI intelligence for governance review

active · owner AI Governance Lead

Staff Drafting Assistant

L1 · Human with assistant

Help staff draft internal correspondence and memos

active · owner Chief Information Officer

Red-Team Probe

L2 · Human-agent team

Adversarial testing of prompts and guardrails on public material

active · owner Security Lead

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.