Service 03 — AI Readiness Assessment
Before you invest further, you need an honest answer to one question.
Is your engineering organization structured in a way that AI can actually work with? Our readiness assessment evaluates five specific dimensions and gives you a clear picture of where you are, where the gaps are, and what to fix first. Most organizations score lower than they expect. That's not a problem — it's a starting point.
What It Is
A diagnostic, not a sales pitch
The AI Readiness Assessment is a time-bounded diagnostic engagement. We run it in three to four weeks, starting from interviews and document review and ending in a written report with specific findings, evidence, and a prioritised list of gaps.
It is not a maturity model comparison against industry benchmarks. It doesn't produce a score that makes you look good or bad relative to peers. It produces a specific account of where your organization is structurally unprepared for the AI investment you're planning to make — or have already made.
What you do with the findings is up to you. The report doesn't commit you to any further engagement with us.
The Five Dimensions
What we evaluate — and why each one matters
Specification clarity
AI systems need requirements that are precise, unambiguous, and testable. Most engineering teams write requirements in natural language that works fine for human developers — who can infer intent, ask clarifying questions, and apply judgment when requirements are incomplete. AI cannot do those things the same way.
We evaluate whether your teams currently write specifications at a level of precision that AI can work with reliably — and identify the gap between what's being specified and what's needed to get consistent output.
Boundary discipline
Every AI deployment involves a boundary between what the AI handles and what a human handles. These boundaries need to be explicit, documented, and enforced — not resolved informally case by case at runtime.
We evaluate whether your organization has clear, agreed contracts at these boundaries — or whether you're operating in an undefined grey zone where neither the AI nor the humans know who is responsible for which decisions. Grey zones are where AI failures compound quietly.
Validation infrastructure
Before you trust AI output — and before you can act on it at scale — you need a reliable way to know when it's correct and when it isn't. Ad hoc spot-checking doesn't scale and doesn't produce consistent results.
We evaluate whether you have systematic mechanisms for validating AI output before it drives decisions — automated tests, sampling protocols, acceptance criteria, or structured review processes. We also look at whether these mechanisms are appropriate for the stakes involved in each use case.
Failure mode awareness
Every AI component has conditions under which it will perform poorly or fail. Some of these are obvious; most aren't. The question isn't whether failures will happen — they will — but whether you'll catch them before they propagate.
We evaluate whether your teams have mapped the specific failure modes of your AI components, and whether you have early warning indicators in place. We distinguish between failure modes that are known and managed from ones that are known but not monitored, and ones that haven't been considered at all.
Decision ownership
When something goes wrong with an AI system — or when it needs to change because the business has changed — someone needs to have the authority and accountability to act. Not just technical ownership (who built it), but operational ownership: who is responsible for what it produces and who can authorize changes to it.
We evaluate whether each of your AI components has a named operational owner with actual authority — or whether ownership is diffuse, contested, or simply absent. Absent ownership is one of the most common reasons AI programs drift rather than improve.
What You Get
A scored assessment across all five dimensions
Each dimension rated with specific evidence from your organization — not an abstract comparison to industry averages, but a concrete account of what we found.
Specific findings, not general observations
Each finding is tied to a specific workflow, team, or system — not a generic statement about "most organizations." You should be able to read each finding and immediately know which part of your operation it refers to.
A prioritised gap list with recommended sequencing
What to fix first, what can wait, and why. Sequenced by impact and by dependency — some gaps need to close before others are worth addressing.
A briefing session with your leadership team
We walk through findings in person or remotely, answer questions, and make sure the report is understood by the people who need to act on it — not just the people who commissioned it.
What You Don't Get
A roadmap for fixing everything
The assessment tells you where the gaps are. The work of designing interventions and sequencing change is a separate engagement. We won't produce a roadmap that looks comprehensive but leaves all the hard decisions unaddressed.
A technology recommendation
The assessment is about organizational structure, not tool selection. Which AI platforms, models, or infrastructure you should use is a different question — and one that shouldn't be answered until the structural foundations are clearer.
Reassurance
If the findings are uncomfortable, we'll say so. The value of the assessment is in its honesty. An assessment that finds no significant gaps and confirms your current direction is a useful thing to have — if it's true. We won't produce one that isn't.
When to Use It
The assessment is most useful at three moments
Before committing significant budget to an AI programme — to understand what structural work is needed first and avoid building on a foundation that won't hold.
Mid-programme, when results aren't coming as expected and the team isn't sure why. The assessment gives you a structured way to find out what's actually limiting performance.
After a previous AI consulting or transformation engagement, to measure where your structural readiness has improved and what still needs attention in the next cycle.
FAQ
Common questions
What happens after the assessment?
You receive the report and the briefing. After that, what you do with it is up to you. Some organisations use it to inform their own internal planning. Others move into a consulting engagement with us to address the gaps. The assessment doesn't obligate you to anything further, and we won't use it as a lever to push you into the next engagement. If you want to discuss what the findings imply for next steps, we're happy to have that conversation — but it's a conversation, not a pitch.
How long does it take?
Three to four weeks from kickoff to report delivery. That includes the interviews, document and codebase review, analysis, and the writing. We don't rush the analysis to hit an arbitrary deadline — the quality of the findings depends on the quality of the evidence gathering.
Who needs to be involved from our side?
We need access to engineering leads, at least one product or operations owner, and ideally someone with cross-cutting visibility across AI initiatives. We're not looking for a steering committee — we're looking for the people who can tell us how things actually work, not how they're documented. Total time commitment from your side is typically four to six hours across the three to four week assessment period.
What if our score is very low?
You found out before investing further — which is the point. Most organisations score lower than they expect, particularly on specification clarity, validation infrastructure, and decision ownership. A low score is a specific starting point: it tells you what to fix, in what order, before you spend more on AI tooling that the organization isn't structured to use effectively.
Can we do this while already running an AI programme?
Yes, and this is often the most useful time to do it. Organisations mid-programme frequently find that the assessment identifies why specific parts of the programme are underperforming — which gives them a clear basis for adjustment rather than continuing to iterate without knowing where the structural constraints are.
Related Services
AI Consulting
The full consulting engagement — diagnoses structural gaps, builds the decision architecture that unblocks your AI investment, and stays through implementation.
Learn more →AI-Native Transformation
If the assessment reveals workflows that are ready to redesign, transformation is the work of actually making those changes at an organizational level.
Learn more →Find out where you actually are before investing further
Three to four weeks. Five dimensions. One honest report. Get in touch to scope the assessment for your organization.
Get Started →