How it works
Confidence at speed comes from architecture, not process. Constellation is corporate governance infrastructure that enforces authority at the moment of action — whether that action is taken by an AI agent, a person using AI tools, or a team making a decision that crosses institutional boundaries. Not in a quarterly review afterwards.
The architecture
Governance is in the critical path. Every action is checked against institutional constraints before it executes — not after.
Model proposes action — any LLM calls a tool via MCP.
Gateway checks authority — constraints evaluated before execution.
Tool executes + trace recorded — provenance captured automatically.
Moments of consequence
Constellation appears only when an action crosses an authority boundary, threshold, timing rule, or irreversibility cliff.
An AI agent about to send an email to a regulator. A team member using Claude to draft a procurement response. A manager approving a financial commitment between meetings. These are moments of consequence — where institutional knowledge needs to be present regardless of who or what is acting. Constellation intervenes here and only here.
Everything else passes through silently.
Constraints checked before acting
Before a consequential action is taken, Constellation checks it against the institution’s goals, permissions, and guardrails — authority rules, spending thresholds, timing restrictions, prohibitions — and returns applicable constraints, violations, and required approvals.
For AI agents, this happens automatically via MCP integration. For humans using AI tools, Constellation is present in the workflow — the same constraints apply whether a response was drafted by Claude or written by hand. The institution’s own rules are consulted before anyone acts.
// Before sending email to all customers
const result = await constellation.check({
action: "send email to 50,000 customers",
domain: "communications"
})
// Response
// { status: "blocked",
// violation: "Exceeds AUTHORITY threshold",
// required: "VP-level approval for comms >1,000 recipients"
// }
Corporate governance infrastructure
Authority, escalation, and acknowledgment are carried by the institution itself — not improvised by people under pressure. Constraints are defined in advance through the Charter, not invented at the moment someone needs to decide.
When a constraint is violated, the institution’s own structure determines who must approve, what authority is required, and how long they have to respond. The governance is in the architecture, not in someone’s memory.
This matters most when AI agents from multiple platforms operate simultaneously. Each platform has its own guardrails, but none governs the institution across all of them. Constellation provides a single set of institutional goals, permissions, and guardrails that apply regardless of which vendor’s agent is acting.
The Proof Layer
Every consequential action produces a governance trace: what happened, who authorised it, what constraints were checked, whether any were violated, and what influenced the decision.
This evidence is produced in real-time, not reconstructed later. When an auditor asks “under whose authority did this happen?” the answer already exists.
We call this governance telemetry — distinct from operational telemetry. Engineering telemetry tells you what the system did. Governance telemetry tells you who decided, on what basis, and with what authority. When something goes wrong, a court or regulator will reconstruct three things: who was responsible, what they knew, and what they did. If your infrastructure doesn’t produce contemporaneous evidence of all three, you are defending yourself from memory.
Who was responsible?
Not who signed the form. Who held institutional accountability for the decision.
What did they know?
What information existed at the time of the decision, recorded contemporaneously.
What did they do?
What action was taken, with what authority, producing what evidence trail.
Standard established in ASIC v Bekier [2026] FCA 196.
The active monitoring standard
Star Entertainment had compliance teams, risk committees, and governance frameworks. The Federal Court still found their governance “dysfunctional.” The distinction is not whether governance exists — it’s whether you can prove it was actively used. Constellation produces this proof automatically: every constraint check, every escalation, every delegation is recorded at the moment it happens.
Human judgment cannot be delegated to AI
The court held that AI can inform decisions but cannot replace personal judgment. Directors must “actively guide and monitor.” Constellation’s accountability chain proves this: every governance trace records the human operator controlling the agent, the agent that acted, and the governance gate that enforced constraints. Human-in-the-loop is not a feature — it’s now a legal requirement.
Regulatory convergence
ASIC v Bekier is not an Australian anomaly. The UK FCA Senior Managers Regime, EU DORA, EU AI Act, and US SEC cybersecurity disclosure rules are all converging on the same standard: executives must demonstrate, with contemporaneous evidence, that governance was active at the point of decision. Governance telemetry is becoming a cross-jurisdictional legal requirement.
The product in action
Real governance telemetry from Constellation. AI agents checked against institutional constraints in real time. Click any trace to see the three-question accountability chain.
Governance Telemetry — Live product preview
Q1: Who was responsible?
Q2: What did they know?
Q3: What did they do?
Every trace answers the three questions a court reconstructs: who was responsible, what they knew, what they did.Standard: ASIC v Bekier [2026] FCA 196.
Governance pulse
Multiple AI agents from different providers, governed by the same institutional constraints. Watch constraint enforcement happen across domains in real time.
Governance Pulse — Simulated day
Communications
Finance
Data & Privacy
Deployments
0 events · 0 passed · 0 blocked · 0 escalated
Every event creates an immutable governance trace
Disagreement as infrastructure
Most governance tools assume compliance. Constellation assumes disagreement.
Any institutional decision can be formally challenged through the Forum. Challenges require standing and evidence. Adjudicators issue rulings. Rulings carry precedent weight that influences future decisions. This is a common law system inside a product.
01
Challenge
File with standing + evidence
02
Review
Adjudicator evaluates evidence
03
Ruling
Decision with reasoning
04
Precedent
Influences future decisions
Progressive trust
AI delegation is not a switch you flip. It is trust earned through evidence.
In Shadow Mode, AI analyses events and recommends actions. The institution decides while Constellation tracks agreement rate per decision category. When accuracy exceeds the threshold with sufficient sample size, delegation becomes available. Delegation can be suspended instantly with full rollback.
You don’t turn on AI. You calibrate trust through evidence.
Confidence you can measure
Not a feeling. A metric.
Every check, escalation, and decision generates data. Constellation measures five dimensions of governance coordination continuously, so you can prove governance is working — not just hope it is.
40%
Faster decisions
100%
Audit compliance
18%
Lower coordination cost
0
Silent authority gaps
Measured across early deployments.
Constellation is intentionally narrow. It does not advise, optimise, or decide. It enforces the institution’s own commitments at the moment they matter.
See where your governance breaks down
Before you can fix coordination, you need to see it clearly. The Governance Health Check measures the five dimensions where institutional knowledge fragments — and shows you exactly where the gaps are.