THE INSTRUMENT

How an AQ score is made.

Practice-anchored questions across three organisational layers, hardened by telemetry, resolved into one live AQ Coordinate. Under 40 minutes per respondent.

A4 · QUESTION BANK A5 · SCORING MATH B · INPUT TEMPLATES
TAKE THE ASSESSMENT → Open a saved report Runs in your browser · answers stay on your device
A1 · THE ASSESSMENT PIPELINE

Six steps, scope to report

SCOPEROSTERRESPONDTRIANGULATERESOLVEREPORT
01 · SCOPE

Scope

The organisation completes the Scoping Input (Template I-1): standard (AQDS/AQOS), units in scope, breadth units declared, telemetry sources available.

02 · ROSTER

Roster

Respondents mapped to the three layers (Template I-2): Executive · Management · Practitioner. Minimums: 3 / 5 / 8 for a High-confidence read; any layer below minimum caps confidence at Medium.

03 · RESPOND

Respond

Each respondent takes an adaptive questionnaire: domain locators first, then depth questions only around their located level. Median session: 25–40 minutes.

04 · TRIANGULATE

Triangulate

Attestation is cross-checked against telemetry bindings and artefact evidence (Templates I-3/I-4). Divergence between said and measured is flagged per domain.

05 · RESOLVE

Resolve

The scoring engine computes domain levels, P and B, the organisational AQ Coordinate and AQ, with confidence and perception-gap analysis.

06 · REPORT

Report

The output template is generated: scorecard, domain profiles, gap analysis, gate plan; advisor review optional before issue. See the sample report ↗

A2 · QUESTION ARCHITECTURE

Every Domain × Level cell is one record

CELL RECORD
Cell ID{domain}·L{n} — e.g. gov·L6 Capability statementOne behavioural sentence of what is true at this level (no metrics, no tools-by-name) Locator contributionThe condition this cell adds to the domain's level-locator ladder Depth questions2–4 questions, 5-point behavioural anchors, each tagged to layers (E/M/P) Gate linksWhich P-gates of the level the cell's confirmation feeds Telemetry bindingsLive signals that can verify or challenge the attestation Breadth counterThe unit over which B is counted for this cell
ANCHOR FAMILIES · 1 → 5
PTSPRACTICEFREQUENCYCOVERAGE
1Not establishedNeverNowhere
2Ad hoc / individualRarelyOne pocket
3Defined, partially followedSometimesSeveral teams
4Managed and reviewedUsuallyMost of the organisation
5Leading practice, continuously verifiedAlwaysEverywhere, verified
LAYER VARIANTS

The same cell asks differently per layer — Executives are asked about mandate and accountability, Management about mechanism and cadence, Practitioners about lived reality. The divergence across layers is the perception-gap input.

A3 · ADAPTIVE FLOW

Locate, confirm, capture

STAGE 1 · 12 QUESTIONS

Domain locators

One per domain. Each locator is a ladder question whose options are condensed level bands; the answer places the domain in a provisional band.

STAGE 2 · BRANCHING

Boundary confirmation

For each domain, the engine asks the depth questions of the provisional level and one level below (cumulative rule) and one above (headroom check). Confirmed = all lower-level conditions hold.

L+1 · headroom checkL · provisionalL−1 · cumulative check
STAGE 3 · CAPTURE

Gate & breadth capture

For the confirmed level: the level's gate checklist (which of the 10 gates are cleared → P) and the breadth counter (coverage across declared units → B).

STOP RULES

A domain that locates at L0–L1 skips depth beyond L2 (no wasted questions); total questions per respondent stay within 60–90 depending on profile; hard cap 40 minutes with save-and-resume.

A4 · THE QUESTION BANK

Published specimens — one locator and one depth question per domain

The full Closed-Tier bank extends each domain to 2–4 depth questions per active level using the same pattern.

wfHybrid Workforcewf·L4 · M
LOCATOR · ALL LAYERS

"Who does the work in your delivery organisation today?"

L0–1People only; any AI use is personal L2–3People with sanctioned AI tools in defined tasks L4–5Redesigned human+AI roles; agents in supervised workflows L6–7Agents own tasks or service areas; humans manage exceptions L8+The workforce mix itself is continuously re-planned by evidence
DEPTH · wf·L4 · MANAGEMENT

"How are roles designed where humans and AI work together?"

1No design; individuals improvise 3Some roles redefined in pilots 5Role architecture covers human+AI pairing across delivery, reviewed on a set cadence
govAgent Governance & Trustgov·L6 · M
LOCATOR · ALL LAYERS

"What may an AI system in your organisation decide without a human?"

L0–1Nothing is defined; nobody could say L2–3Usage policy exists; decisions stay human L4–5Defined assistive scope; human approval on actions L6–7Written autonomy boundaries per agent, with escalation and audit L8+Boundaries are tested, telemetry-verified, and adjusted under governance
DEPTH · gov·L6 · MANAGEMENT

"When an agent acts outside its boundary, what happens next?"

1We'd find out eventually 3It's noticed and handled informally 5Automatic halt, alert, audited post-mortem within a defined SLA
ctxContext & Intelligence Engineeringctx·L5 · P
LOCATOR · ALL LAYERS

"When an AI system needs organisational knowledge, where does it come from?"

L0–1Individuals paste what they have L2–3Shared prompt/document stores L4–5Curated context pipelines with ownership and freshness rules L6–7Governed knowledge graph/retrieval serving agents with access control L8+Context quality is measured and self-maintaining
DEPTH · ctx·L5 · PRACTITIONER

"How fresh and reliable is the context your AI tools draw on?"

1Often wrong or stale; we double-check everything 3Mostly usable; gaps are known 5Owned, dated, access-controlled; staleness is flagged before it bites
mpModel & Platform Operationsmp·L6 · M
LOCATOR · ALL LAYERS

"How are AI models chosen and run for your work?"

L0–1Whatever an individual uses L2–3A sanctioned default model/tool L4–5A managed portfolio; cost and fit considered per task L6–7Tiered routing by task with evaluation harnesses and cost telemetry L8+Routing and portfolio adapt automatically within governance
DEPTH · mp·L6 · MANAGEMENT

"How do you know a model change won't degrade a running service?"

1We don't; we'd hear complaints 3Spot checks before switching 5Evaluation harness gates every change; regression blocks promotion
valValue & Performance Intelligenceval·L4 · E
LOCATOR · ALL LAYERS

"How is the contribution of AI to outcomes measured?"

L0–1It isn't L2–3Anecdotes and satisfaction L4–5Throughput/quality deltas measured on defined work L6–7Live attribution of human vs agent contribution; cost-to-serve tracked L8+Value telemetry drives re-baselining automatically
DEPTH · val·L4 · EXECUTIVE

"Could you defend the AI contribution number in a board meeting?"

1There is no number 3Directional estimates 5Measured, attributed, trended — and priced into engagements
relRelationship & Ecosystemrel·L5 · M
LOCATOR · ALL LAYERS

"How do clients and partners experience your hybrid delivery?"

L0–1They don't know AI is involved L2–3Disclosed where asked L4–5Agreed and visible in ways of working L6–7Interfaces (including agent-to-agent) defined with key partners L8+Ecosystem operating agreements govern cross-boundary autonomy
DEPTH · rel·L5 · MANAGEMENT

"When your delivery includes agents, how is that agreed with the client?"

1It isn't discussed 3Mentioned informally 5Explicit in the engagement model: scope, oversight, and contacts for agent-related issues
ocOutcome Contractingoc·L6 · E
LOCATOR · ALL LAYERS

"What do your contracts say about AI and autonomy?"

L0–1Nothing L2–3Generic tool clauses L4–5AI-assisted delivery acknowledged; outputs warranted by humans L6–7Autonomy scope, AI service-level objectives, and accountability for agent actions are contractual objects L8+Outcome-based commercials re-price as autonomy deepens
DEPTH · oc·L6 · EXECUTIVE

"If an agent's action caused a client loss tomorrow, is accountability already written down?"

1No 3Arguably, in general clauses 5Explicit allocation: scope, liability, remedy, and evidence obligations
scSolution Compositionsc·L6 · P
LOCATOR · ALL LAYERS

"How are AI/agentic solutions built and approved for use?"

L0–1Ad hoc builds go straight to use L2–3Peer review before use L4–5Defined build method with pre-deployment testing L6–7Simulation + safety case required before autonomy is granted L8+Composition itself is agent-assisted under the same gates
DEPTH · sc·L6 · PRACTITIONER

"Before an agent gets real autonomy, what does it have to pass?"

1Nothing formal 3A demo and a sign-off 5Simulated scenarios, boundary tests, and a written safety case someone owns
aoAutonomous Operationsao·L7 · M
LOCATOR · ALL LAYERS

"Who runs your services day to day?"

L0–1People, with tickets L2–3People, with AI-assisted diagnostics L4–5Automations handle known-good runbooks; humans supervise L6–7Agents run service areas; humans manage exceptions; self-healing is normal L8+Operations improve themselves within governed bounds
DEPTH · ao·L7 · MANAGEMENT

"When something breaks at 3 a.m., what usually happens before a human is involved?"

1Nothing; a human is paged first 3Diagnostics gathered automatically 5Detection, remediation attempt, verification, and an auditable record — humans see the summary at 9
evContinuous Evolutionev·L6 · M
LOCATOR · ALL LAYERS

"How do your AI capabilities improve once live?"

L0–1They don't; they age L2–3Occasional manual updates L4–5Scheduled reviews with measured improvements L6–7Instrumented learning loops with governed promotion L8+Agents improve agents; value re-bases automatically
DEPTH · ev·L6 · MANAGEMENT

"Describe the path from 'the agent could do better' to 'the improved agent is live'."

1There is no path 3Someone raises it; eventually fixed 5Loop is instrumented: signal → candidate → evaluation → governed promotion, on a cadence
arAssurance & Resiliencear·L6 · P
LOCATOR · ALL LAYERS

"How prepared are you for AI-specific attacks and failures?"

L0–1Not considered L2–3General security covers it, we assume L4–5AI-specific threats assessed; controls defined L6–7Tested defences (injection, poisoning, agent compromise) with autonomy-fallback-to-human drills L8+Continuous adversarial testing; resilience is measured
DEPTH · ar·L6 · PRACTITIONER

"If autonomy had to be switched off in one service right now, what would happen?"

1Chaos; nobody knows 3We'd cope with disruption 5Documented fallback runs on drills; service degrades gracefully to human operation
trCapability Transfertr·L5 · M
LOCATOR · ALL LAYERS

"If an engagement ended next quarter, what could the client take with them?"

L0–1Documents, maybe L2–3Documentation and handover sessions L4–5Structured knowledge transfer with client capability uplift L6–7Context, agents, and knowledge are portable by design; exit is rehearsed L8+Transfer readiness is continuously verified as part of delivery
DEPTH · tr·L5 · MANAGEMENT

"How early in an engagement is exit portability designed?"

1It isn't; we'd scramble 3Addressed near the end 5From day one: portability requirements in the solution design and checked at milestones
A5 · SCORING MATH

From cells to one Coordinate

PER DOMAIN

DomainLevel(d) = highest L where the locator band and all depth confirmations at L and below hold (cumulative rule).

P(d) = gates cleared at DomainLevel ÷ 10 × 10 → 0–10.

B(d) = breadth counter coverage across declared units → 0–10.

ORGANISATIONAL AQ COORDINATE

L(org) = median of the twelve DomainLevels.

P(org) = mean P of domains at L(org) · B(org) = mean B of domains at L(org).

AQ = (L × 10) + (0.6·P + 0.4·B) × (10/11)
THE TRUST GATE

L(org) ≤ min(gov, ar) + 1 — autonomy may never run more than one level ahead of governance and assurance.

LAYERS & PERCEPTION

Attestation resolves per layer (E/M/P) unweighted; the published score uses the triangulated resolution (attestation × telemetry × artefacts), while the layer split and its gaps are reported, never averaged away.

CONFIDENCE

High (all layer minimums met, telemetry connected, low variance) · Medium (one shortfall) · Low (two or more) — always printed beside the AQ.

PART B · INPUT TEMPLATES

What the instrument asks for

I-1 · SCOPING INPUT
Organisation Standard (AQDS / AQOS) Units in scope (list; these are the breadth units) Headcount · engagements in scope Telemetry sources available: agent action logs / oversight events / routing & cost / evaluation results / incident records Assessment window
I-2 · RESPONDENT ROSTER

Per respondent: name (held encrypted) · unit · layer (Executive / Management / Practitioner) · channel (email / link / messaging / voice).

E 6 / 3 ✓ M 12 / 5 ✓ P 20 / 8 ✓

Individual responses are encrypted; only aggregates are ever displayed.

I-3 · TELEMETRY INTAKE

Per connected source: signal name · cell bindings (e.g., boundary-breach rate → gov·L6) · period · collection method (API / export).

A "no telemetry" path is supported — it caps certification standing until connected.

I-4 · ARTEFACT REGISTER

Per artefact: type (autonomy charter / safety case / policy / contract clause / audit trail) · cells evidenced · date · reviewer.

AI-first review, human verification for certification-grade standing.

See what the instrument produces.

Read the sample report

The specimens on this page are the published subset of the AutonowmyQ instrument (v1.0). The full question bank, gate definitions, telemetry bindings and calibration data are Closed-Tier and available only under licence. AutonowmyQ™, AQ™ (Autonowmy Quotient), AQDS™, AQOS™, AQCS™ and AQ Certified™ are marks of Autonowmy Technology Pvt Ltd; registrations pending.