What does an AI readiness assessment include?
The short answer
An AI readiness assessment looks at five things: how your work actually flows, what data you hold and where it lives, which tools and licenses you already pay for, what your confidentiality and risk constraints allow, and who on the team would use this. It ends with a short list of tasks worth doing first, each with an hours estimate attached.
Automation & AI · Reviewed by Bracey Skyway Partners, Buffalo, NY
What gets examined
- Process map
- How work moves through the business today, including the steps that live in someone's head and the ones that only happen when a particular person is in the office.
- Data and documents
- What you hold, what condition it is in, and whether a model could be grounded in it. Most of the value sits in proposals, past emails, pricing, and call notes rather than in a database.
- Tools and licenses
- What you already pay for. A surprising share of readiness work ends with turning on capability that was included in a subscription nobody had read to the end.
- Confidentiality and risk
- Where client data, health information, or contractual terms rule out a given tool, and what has to be true before anything touches a customer.
- People and appetite
- Who would actually use this, who has tried already, and who quietly believes it is a fad. That last group decides whether adoption survives the first month.
- A ranked shortlist
- Three to five candidate tasks, each with the hours it consumes today, what changes, what it would take to build, and what could go wrong.
What it is not
It is not a maturity score out of five, and it is not a slide deck about what the technology could theoretically do for your sector. Those read well and change nothing, because they never name a task, an owner, or a number.
An assessment that does not end with something you could start on Monday has failed, regardless of how thorough it looked.
How long it takes
Two to four weeks for most companies. The bulk of that is sitting with the people doing the work and watching what they actually do, which is reliably different from the documented process.
The output is short: the map, the shortlist, the constraints, and what we would build first. Long enough to act on, short enough that the leadership team reads all of it.
What usually comes out of it
Two patterns come up again and again. The first is that the highest-value task is not the one leadership expected, because the expensive work is buried in admin rather than in the parts of the business people talk about.
The second is that a third of the shortlist turns out to be rule-shaped, and should be automated rather than modeled. That is a cheaper answer, so it is the one worth finding early.
Let’s put a plan — and a build — behind your next move.
Tell us what’s slowing the business down, and we’ll tell you what we would build first and what it would take.