Example report
Organisation Deep Dive Assessment
Department Deep Dive: Customer Operations (example)
A useful basis exists but material confirmation or improvement remains. Identify specific gaps and owners.
The score reflects the declared maturity of each position and the basis on which the answer is given. It is not a certification, a valuation, an implementation approval or a guarantee of suitability, and it is not a measure of process quality or organisational maturity.
4 of 11 relevant gates need a condition. A favourable average cannot clear a gate. Treat each condition below as a blocker, or as a need for specialist confirmation, before relying on this result.
Confidence profile
- Scored questions
- 63 of 92
- Completion
- 100%
- Unknown
- 6
- Disputed
- 6
- Not applicable
- 0
- Gates
- 4 of 11 need a condition
- Assessment type
- Deep Dive completed online (not yet facilitated)
- Scope assessed
- Department: Customer Operations (example)
- Assessment date
- 2026-10-01
Conditional findings
Reported separately. These scores are not part of the overall figure.
Gates and conditions
- Condition Does every non-development agent have a unique, managed identity attributable.... Condition needed: Below 60% confidence.
- Condition Are permitted, conditional and prohibited decisions and actions explicitly.... Condition needed: Below 60% confidence.
- Condition Can the agent recognise insufficient information, uncertainty, conflict,.... Condition needed: Unknown.
- Condition Are authority boundaries, human oversight, containment, rollback and recovery.... Condition needed: Below 60% confidence.
- Clear Are agent-specific threats assessed, including goal hijacking, prompt.... No condition.
- Clear Are high-impact or irreversible tool calls subject to independent policy.... No condition.
- Clear Is each agent's purpose expressed as a bounded, testable objective with.... No condition.
- Clear Are human approval, monitoring, intervention, override and escalation points.... No condition.
- Clear Are human reviewers given the information, time, competence and authority.... No condition.
- Clear Are combined and cumulative actions controlled so that multiple agents cannot.... No condition.
- Clear Can an agent, permission, tool or action type be immediately restricted or.... No condition.
Information that would improve confidence
Information is currently unknown. Identify an accountable source and clarify the position.
Governance Operating Model and Accountability
-
governance_5 Is independent review or challenge required for higher-risk or higher-impact or...0%
Tests whether decisions and accountability can be exercised consistently.
Process and Operating-Model Readiness
-
process_5 Can dependencies between processes, information, systems, controls, policies,...0%
Tests whether the organisation understands and can redesign how work happens.
Technology Architecture and Integration
-
architecture_5 Are models, prompts or instructions, tools, policies, knowledge sources and...0%
Tests whether technology can be integrated, operated and changed sustainably.
-
authority_5 Can the agent recognise insufficient information, uncertainty, conflict,...0%
Tests whether agent authority remains bounded by human accountability.
Operational Control, Monitoring, Incident Response and Resilience
-
operations_4 Is there a tested incident plan covering detection, escalation, containment,...0%
Tests whether deployed change can be observed, contained and improved.
Workforce, Organisation and Change
-
workforce_4 Are people trained in the context of their role to use, supervise, challenge,...0%
Tests whether people can adopt, supervise and sustain the operating model.
Relevant participants disagree. Reconcile the different views before relying on this conclusion.
Governance Operating Model and Accountability
-
governance_3 Is accountability for business outcomes distinguished from technical, solution,...13%
Tests whether decisions and accountability can be exercised consistently.
Process and Operating-Model Readiness
-
process_3 Can actual working practices, exceptions, workarounds, rework, queues and...13%
Tests whether the organisation understands and can redesign how work happens.
Technology Architecture and Integration
-
architecture_3 Are integrations controlled, supportable and observable across cloud, SaaS,...13%
Tests whether technology can be integrated, operated and changed sustainably.
-
authority_3 Is the proposed level of autonomy necessary to create the intended value,...13%
Tests whether agent authority remains bounded by human accountability.
Operational Control, Monitoring, Incident Response and Resilience
-
operations_2 Are thresholds and alerts defined for performance deterioration, unusual...13%
Tests whether deployed change can be observed, contained and improved.
Workforce, Organisation and Change
-
workforce_2 Are changes to tasks, decisions, accountability, workload, skills, controls and...13%
Tests whether people can adopt, supervise and sustain the operating model.
Confidence is limited. Establish the position, owner and supporting internal basis.
Governance Operating Model and Accountability
-
governance_7 Does the board or governing body receive reporting that supports specific...13%
Tests whether decisions and accountability can be exercised consistently.
Process and Operating-Model Readiness
-
process_7 Can the organisation prevent an initiative from digitising or automating an...13%
Tests whether the organisation understands and can redesign how work happens.
Security, Agent Identity and Access
-
security_1 Does every non-development agent have a unique, managed identity attributable...13%
Tests whether identity, access and technology-specific threats can be controlled.
-
authority_7 Where an agent delegates to another agent, do the original purpose,...13%
Tests whether agent authority remains bounded by human accountability.
Operational Control, Monitoring, Incident Response and Resilience
-
operations_6 Are incidents, near misses, human overrides, complaints and recurring...13%
Tests whether deployed change can be observed, contained and improved.
Workforce, Organisation and Change
-
workforce_6 Is operational capacity funded for oversight, exceptions, incident response,...13%
Tests whether people can adopt, supervise and sustain the operating model.
Moderate confidence 34 items
Confidence is moderate. Clarify gaps and obtain accountable-owner confirmation.
Strategy, Purpose and Value
-
strategy_6 Is there a phased roadmap linking increased agent authority to demonstrated...43%
Tests whether change is directed by business purpose rather than technology enthusiasm.
-
strategy_5 Is there a named executive sponsor with authority over funding,...53%
Tests whether change is directed by business purpose rather than technology enthusiasm.
-
strategy_1 Is there a documented strategy for process and operating-model improvement...64%
Tests whether change is directed by business purpose rather than technology enthusiasm.
Use-Case Portfolio and Risk Classification
-
portfolio_6 Is the maximum credible impact of a single action, a sequence of actions and...43%
Tests whether opportunities and risks can be governed as a portfolio.
-
portfolio_5 Are prohibited, restricted and higher-risk uses identified before technology...53%
Tests whether opportunities and risks can be governed as a portfolio.
-
portfolio_1 Is there a current inventory of proposed, experimental, deployed, suspended and...64%
Tests whether opportunities and risks can be governed as a portfolio.
Data, Knowledge, Context and Memory
-
data_6 If an AI agent retains information between tasks or conversations, are there...43%
Tests whether information can support dependable and lawful operation.
-
data_5 Are controls defined for missing, stale, conflicting, poisoned, untrusted or...53%
Tests whether information can support dependable and lawful operation.
-
data_1 Can the organisation identify the information a proposed solution may read,...64%
Tests whether information can support dependable and lawful operation.
Security, Agent Identity and Access
-
security_7 Are agent-to-agent and agent-to-tool communications authenticated, authorised,...43%
Tests whether identity, access and technology-specific threats can be controlled.
-
security_6 Can agents distinguish authoritative instructions from untrusted content such...53%
Tests whether identity, access and technology-specific threats can be controlled.
-
security_2 Are permissions restricted by task, system, data, action, transaction, duration...64%
Tests whether identity, access and technology-specific threats can be controlled.
Testing, Evaluation and Independent Assurance
-
assurance_5 Are evaluation data, test methods, limitations, failures and residual risks...43%
Tests whether performance, control and limitations can be demonstrated before reliance.
-
assurance_4 Are authority boundaries, human oversight, containment, rollback and recovery...53%
Tests whether performance, control and limitations can be demonstrated before reliance.
-
assurance_7 Are regression tests repeated after material changes to processes, models,...64%
Tests whether performance, control and limitations can be demonstrated before reliance.
Legal, Regulatory, Ethical and Affected-Person Rights
-
legal_5 Are bias, discrimination, accessibility and differential outcomes evaluated...43%
Tests whether obligations and affected-person rights are translated into practice.
-
legal_4 Can people affected by the process obtain meaningful human review, challenge,...53%
Tests whether obligations and affected-person rights are translated into practice.
-
legal_7 Is regulatory and policy change monitored and translated into updated process...64%
Tests whether obligations and affected-person rights are translated into practice.
Third Parties, Economics and Lifecycle Roadmap
-
ecosystem_5 Are financial consumption limits and controls in place to prevent uncontrolled...43%
Tests whether suppliers, economics and lifecycle dependencies remain governed.
-
ecosystem_4 Is total cost measured per successful business outcome, including integration,...53%
Tests whether suppliers, economics and lifecycle dependencies remain governed.
-
ecosystem_7 Does the lifecycle roadmap include decision gates for experimentation,...64%
Tests whether suppliers, economics and lifecycle dependencies remain governed.
Governance Operating Model and Accountability
-
governance_2 Are decision rights documented for approving investigation, pilots, production...53%
Tests whether decisions and accountability can be exercised consistently.
-
governance_1 Is one accountable executive identified for the organisation's overall agentic...64%
Tests whether decisions and accountability can be exercised consistently.
Process and Operating-Model Readiness
-
process_2 Do priority processes have named business owners, agreed boundaries, intended...53%
Tests whether the organisation understands and can redesign how work happens.
-
process_1 Can the organisation identify the end-to-end processes and value streams in...64%
Tests whether the organisation understands and can redesign how work happens.
Technology Architecture and Integration
-
architecture_2 Can required capacity, latency, availability and continuity be provided within...53%
Tests whether technology can be integrated, operated and changed sustainably.
-
architecture_1 Is there an approved reference architecture covering agent platforms, models,...64%
Tests whether technology can be integrated, operated and changed sustainably.
-
authority_2 Are permitted, conditional and prohibited decisions and actions explicitly...53%
Tests whether agent authority remains bounded by human accountability.
-
authority_1 Is each agent's purpose expressed as a bounded, testable objective with...64%
Tests whether agent authority remains bounded by human accountability.
-
authority_8 Are combined and cumulative actions controlled so that multiple agents cannot...64%
Tests whether agent authority remains bounded by human accountability.
Operational Control, Monitoring, Incident Response and Resilience
-
operations_1 Can material inputs, context, tool calls, approvals, decisions, actions,...53%
Tests whether deployed change can be observed, contained and improved.
-
operations_7 Are operational controls and recovery arrangements retested at a frequency...64%
Tests whether deployed change can be observed, contained and improved.
Workforce, Organisation and Change
-
workforce_1 Are affected employees and representative users involved in process redesign,...53%
Tests whether people can adopt, supervise and sustain the operating model.
-
workforce_7 Are customer and employee experience outcomes measured alongside productivity...64%
Tests whether people can adopt, supervise and sustain the operating model.
Prioritised actions
Owner for every action: To be confirmed.
- High Confidence improvement. Identify an accountable source and establish the position:
- Can the agent recognise insufficient information, uncertainty, conflict,... authority_5, gate OD-AUTHORITY_5
- Is independent review or challenge required for higher-risk or higher-impact or... governance_5
- Can dependencies between processes, information, systems, controls, policies,... process_5
- Are models, prompts or instructions, tools, policies, knowledge sources and... architecture_5
- Is there a tested incident plan covering detection, escalation, containment,... operations_4
- Are people trained in the context of their role to use, supervise, challenge,... workforce_4
- High Capability improvement. Establish or strengthen:
- Does every non-development agent have a unique, managed identity attributable... security_1, gate OD-SECURITY_1
- Does the board or governing body receive reporting that supports specific... governance_7
- High Confidence improvement. Obtain accountable-owner confirmation of:
- Are permitted, conditional and prohibited decisions and actions explicitly... authority_2, gate OD-AUTHORITY_2
- Are authority boundaries, human oversight, containment, rollback and recovery... assurance_4, gate OD-ASSURANCE_4
- High Capability improvement. Reconcile the differing views:
- Is accountability for business outcomes distinguished from technical, solution,... governance_3
- Can actual working practices, exceptions, workarounds, rework, queues and... process_3
Assumptions and limitations
This result rests on the answers and the basis you declared. IGX has not independently validated them, and no confidential evidence was requested or reviewed. It was completed online without a facilitator, so it should not be compared directly with a facilitated Deep Dive. A self-assessment may be compared with another only where the scope, the respondent basis and the completion rate are alike.
Recommended engagement
Work through the prioritised actions with their owners, resolve every gate condition, then reassess.
Get your own result
It takes a few minutes. Your result appears on screen when you finish, and a private link to it is emailed to you.