Process Deep Dive Assessment
Challenge the evidence before you decide the operating mode.
In about 55 minutes, test the evidence behind your view of a process with the people who know it, and see which operating approach it supports. The guidance below the assessment helps you give answers you can defend.
Find out how much of what you know you can rely on.
- What your people know about how the process runs, written down and checked.
- Which operating approach the evidence supports: improve, simplify, retain, defer or stop.
- Where Agentic AI would need controls and authority before it goes near the process.
- The conditions for a decision you can defend.
Who else should see this?
Answer from where you sit. A first view from one person is useful, and it carries more weight once colleagues who own other areas have checked it. Ask them to take the same assessment, then compare.
Your answers carry most weight when
- You are accountable for the area.
- You know how the work is actually done.
- You can say who owns governance, technology and data decisions.
- You can tell fact from assumption.
- You can reach people who can check your answers.
Where you are less sure
Say so. "Unknown" is a useful answer, and the result shows you where to look first.
Assess one process, then only the modes that are credible.
Answer against the process as it stands today, and record the basis behind each answer. Where two answers seem to apply, choose the lower one unless the higher state is confirmed. Confirm existence, ownership and approval status only, and do not enter budgets, personal data or credentials. The result appears on screen as soon as you finish, and a private link to it is emailed to you.
Loading the assessment.
How useful was this assessment?
Your feedback is critical to improving this assessment, and every comment is read by the team. Rate it, add a comment if you can, and we will email your result link straight after.
Choose a rating above to send your rating and comments. Skipping does not send them.
One process, its evidence and its operating options.
This is not a general judgement about technology. The common core tests the process scope and truth, its fitness and the case for change, demand and variation, the comparative operating fit, information and technology, consequence and control, people and operational readiness, and the assurance behind a decision.
Neither track assumes progression towards more automation or Agentic AI. Retaining or improving the current arrangement is a legitimate outcome.
Begin with a process clear enough to assess.
A Deep Dive suits a process where consequences are material, views conflict, confidence is limited or the operating options are disputed. Identify the process owner, the executive sponsor, the intended outcome, the systems and information involved, the people affected and the constraints that apply.
Not sure where your confidence stands?
Run the free Process Readiness Assessment first. It shows how far your understanding of the process can be relied on.
Assess a processOwnership or governance unresolved?
Run the Organisation Readiness Assessment with a multidisciplinary team to establish the wider context.
Start the assessmentAn average is not allowed to hide a gate that has failed.
- Confidence is separate from direction. The report explains what is understood and how confidently before it says anything about direction. Direction is a judgement made from the pattern of answers, gates and modules. It is never the mode with the highest score.
- Every answer carries its basis. For each question you give a response and how you know: approved, owner confirmed, team consensus, an individual view or disputed. An unknown scores zero and stays visible.
- Gates constrain the result. Gates cover the common core and the modules you activate. A gated concern cannot be averaged away by strengths elsewhere.
- Modules apply only when they are credible. Human-led, workflow, task automation, technology assistance, supervised Agentic AI and governed autonomy are chosen after the common core. A module that is relevant but left incomplete lowers confidence. One that is not relevant is recorded as not applicable, with your reason.
- The calculation is deterministic. The score is produced by fixed rules. A language model does not decide the result. Every result records the version of the questions and scoring method used.
Activate a module only when it is a credible option.
You choose the modules after the common core. Each one has a test for when it applies and a reason it should not be activated.
- Human-led. Activate when: Human execution remains a credible target or retained component. Not merely because: The process currently uses people.
- Workflow. Activate when: The problem concerns coordination, routing, hand-offs, approvals or visibility. Not merely because: A workflow product is available.
- Task automation. Activate when: Stable, repeatable tasks are credible candidates for rules, integration or RPA. Not merely because: There is pressure to automate.
- Technology assistance. Activate when: People should retain decision or execution control with tool support. Not merely because: AI features exist in a platform.
- Supervised Agentic AI. Activate when: Adaptive interpretation or action appears necessary and simpler modes are insufficient. Not merely because: Agentic AI is fashionable or technically possible.
- Governed autonomy. Activate when: Adaptive interpretation or action appears necessary and simpler modes are insufficient. Not merely because: Agentic AI is fashionable or technically possible.
Five views of confidence.
A score is not a certificate. A band is not an approval.
- 1
Confidence score and band
The average confidence across the common core, in one of four bands. It reflects how established each position is and the basis on which the answer is given. It is not a measure of process quality.
- 2
Domain and module scores
A score for each of the eight core domains, and a separate score for each operating-mode module you completed. Module scores are never blended into the overall figure.
- 3
What each step needs to succeed
Where a relevant safeguard is not yet established, unknown, disputed or below 60 per cent confidence, the result shows the condition to meet or the specialist confirmation needed. A safeguard is never a scoring bonus, and a strong average cannot replace it.
- 4
Information that would improve confidence
The unknowns, disagreements and weakly supported answers, each with why it matters and the internal basis that would confirm it.
- 5
Prioritised actions
Actions split into confidence improvement and capability improvement, traced to the questions and gates that created them, ready to carry into the Readiness Roadmap.
The assessment ends with owned action, not a PDF score.
Every material finding traces to a question, a gate or an information gap and becomes an action with an owner, a dependency, the evidence required to close it, a decision gate and a reassessment trigger. That is the Readiness Roadmap.
Practical questions.
Do I have to complete every operating-mode module?
No. Answer the common questions, then only the operating-mode modules that are credible for the process. A module that does not apply is recorded as not relevant, with the reason.
Should I complete the Process Readiness Assessment first?
The Deep Dive normally follows where the Process Readiness Assessment indicates deeper work is warranted. The Readiness Assessment is free, takes about 20 minutes and shows how far your understanding of the process can be relied on. The Deep Dive then tests it with evidence.
Does the result recommend automation?
No. It shows how much confidence each answer deserves, which gates need a condition, and which operating modes are credible for the process. A simpler, human-led arrangement, or a decision to retain, defer or stop, remains a valid outcome.
The Process Deep Dive Assessment provides an indicative decision-support baseline. It does not constitute certification, independent assurance, financial validation, legal, regulatory, security or technical advice, or approval to implement a change. IGX does not independently validate the information unless expressly stated. A competent accountable body must make the decision using appropriate specialist evidence for the process, sector, affected people and jurisdictions involved.