How I work
OAR — Operational AI Readiness
A process-level diagnostic that identifies where AI creates real operational value now, where it does not, and what the first move should be.
Built for leaders in operationally complex businesses — transport, logistics, manufacturing, public sector. Not a generic AI maturity assessment. Not a technology shopping list. A map of where your real operation is ready, where it is not, and what to do about it.
What OAR evaluates
Five dimensions
Each process in your operation gets scored across these five dimensions. The combined signal tells you whether AI belongs here at all — and if so, what shape it should take.
Repetitiveness
How repeatable is the process?
AI works best where the same shape of decision gets made hundreds of times a week. If every case is a snowflake, automation will create more work than it removes. The first thing I map is the actual distribution — not what the process document says, but what the team does on a Tuesday afternoon.
Data quality
Is data available in a usable format?
Most operational data exists somewhere. That is not the same as being usable. I look for data that is current, complete, and structured enough that a model can learn from it without heroic cleanup. When the answer is no, the first project is almost never AI — it is the data groundwork that makes AI possible later.
Decision structure
Formalized logic or tacit knowledge?
Some decisions are written down in SOPs. Others live in the head of the one person who has been there fifteen years. Both can be automated, but they require different approaches and different risk tolerance. Mistaking tacit knowledge for formalized logic is how most AI pilots fail quietly.
Staff readiness
Time, skills, and mandate to act?
A well-designed AI tool in the hands of people who have no time to use it is a shelf ornament. I check whether the team closest to the work has the bandwidth, the baseline skills, and — critically — the mandate to change how they do things. If any of those three are missing, fix them first.
Incentive structure
Does the organization reward the right behavior?
This is the dimension that gets skipped and it is the one that decides whether any of the other four matter. If the incentives reward volume and the AI tool improves quality, the tool will be ignored. If the incentives reward speed and the tool adds a safety check, it will be bypassed. Look at what people are measured on before you decide what to build.
What you get
Four concrete deliverables
No strategy deck. No 40-page report no one reads. Four outputs you can act on the same week.
- 01
Process overview
Green / yellow / red signal per workflow, based on all five dimensions.
- 02
Priority list
The 3–5 processes worth moving first, ranked by realistic value and effort.
- 03
Action plan
What to test, what to prepare, what to avoid — concrete next steps, not a strategy deck.
- 04
Anti-list
Where AI is not the answer right now, and what to do instead.
The first conversation is free and takes thirty minutes. You leave with a clearer picture of where AI fits in your operation — and where it does not.