We start with the work you want to improve. You choose the scope, which can include any part of your organisation.
What we do together
01
Agree what to look at
We agree the work to review, the decision you face, who will make it and when. We include people who do or support the work.
02
Find where work is slow
We look for recurring delays and the work behind your AI ideas. We choose one or two areas to examine more closely.
03
Work out the potential benefit
I compare AI with the current approach and the best non-AI option. We estimate the benefit using what you can measure, including errors, checking and support time.
04
Rate readiness
I rate your ideas, organisation and technology, with reasons for each finding. If evidence is unclear, I ask more questions or draw on professional experience. I explain my reasoning and uncertainty.
05
Recommend next steps
You check the draft for factual errors. I then give you my recommendation, brief your decision-makers and set out the six-month plan.
What I base the assessment on
I check important claims against current documents, measures and conversations. I distinguish what I found from estimates, assumptions and professional judgement.
A written policy does not show whether people use it. A missing document does not show that the practice is missing.
I use descriptions of roles and summary information. Please leave out personal records, names, passwords and confidential system details.
How strong is the evidence?
Confidence describes the evidence behind a finding. Where it is unclear, I ask more questions or use professional judgement, explaining what I infer and why.
What the confidence labels mean
Strong
Current, direct evidence covers the finding, and the method fits the question. No important contradiction remains unresolved.
Moderate
Evidence supports the finding, but leaves an important gap, assumption or limit.
Limited
Evidence is incomplete, indirect or contradictory.
The levels describe how established current practice is in the work we review. Optimised is not a target for every organisation.
Read the descriptions for all three areas
AI ideas and value
The problem an idea would solve, its expected value, how you choose priorities and how you measure results.
Unplanned
Ideas lack a clear task, intended user or expected result. There is no agreed way to choose between them.
Ad hoc
Teams have ideas and possible benefits, but use different evidence and measures to compare them.
Emergent
The main ideas address defined problems, with expected benefits and some measures. Important assumptions or comparisons still need testing.
Standardised
Teams use shared criteria to compare ideas and alternatives. They measure agreed outcomes when deciding what to do next.
Optimised
Results from completed work change priorities. Evidence about benefits, costs, quality and user impact shows which ideas to improve, change or stop.
Organisational readiness
Who is responsible, the guidance and skills people need, and how the organisation handles risk and change.
Unplanned
The people involved cannot confirm who is responsible for the work, what guidance applies, who checks results or who provides support.
Ad hoc
Some teams have owners and informal guidance, but responsibility and support vary across the organisation.
Emergent
Roles, guidance, skills and support for the main ideas are taking shape. Some teams still lack them or do not use them.
Standardised
People across the reviewed work consistently use agreed approval, review, training and feedback practices. They know their responsibilities and can raise concerns and correct errors.
Optimised
Outcomes, feedback and problems lead to changes in responsibilities, guidance, training and work practices. Those changes are visible in day-to-day work.
Technology readiness
The information, tools, access, connections, purchasing and support the proposed work would need.
Unplanned
The organisation has not identified the information, systems, access, connections and support needed for the work.
Ad hoc
Tools or small trials exist, but teams decide access, connections, purchasing and support separately for each case.
Emergent
The likely systems, information and access needs are clear. Some questions about connections, purchasing or running the service remain open.
Standardised
The team consistently manages approved systems, information rights, access, connections, monitoring and support. It has a backup approach if the service fails.
Optimised
Evidence about reliability, performance, cost, access and incidents leads to regular improvements in technical choices and how the service runs.