Skip to content

Decide before you build

AI Readiness Assessment

Find out where AI can create measurable value, what would block delivery and which first project deserves investment. The assessment is grounded in real workflows and representative examples, not a generic maturity score.

Discuss this engagement

The decision this supports

Which AI opportunity should we fund first, and what must be true for it to work?

For leadership, operations and technology teams that have several AI ideas but no shared evidence for comparing them. It also helps when an earlier pilot looked promising but never became dependable operational software.

Signals this may fit

  • Different teams are proposing unrelated AI projects with no common decision criteria.
  • Data exists, but access, ownership, quality or privacy constraints are unclear.
  • A vendor demonstration worked while the real workflow and exception cases remain untested.
  • Leadership needs a defendable build, buy, change-process or stop recommendation.

Start with evidence

Bring representative, non-sensitive examples from the current workflow, including routine work and exceptions. We use them to make the scope and acceptance criteria concrete before expanding the build.

What you receive

Workflow and constraint map
A concise view of the current task, handoffs, exceptions, systems, owners and cost of the existing problem.
Prioritized opportunity shortlist
Candidate use cases compared by likely value, feasibility, adoption effort, evidence needed and consequence of failure.
First-pilot decision
A bounded test plan with representative examples, acceptance criteria, dependencies and explicit go, change or stop conditions.

How the engagement works

  1. 01

    Frame

    Choose the decisions and operational outcomes the assessment must support.

  2. 02

    Inspect

    Review workflows, systems, data access, difficult examples and existing controls with the people who own the work.

  3. 03

    Test

    Challenge the riskiest assumption with representative inputs instead of relying on a polished demonstration.

  4. 04

    Decide

    Compare AI, standard software, process improvement and postponement using the same evidence.

What shapes scope and cost

  • How many workflows, business units and stakeholder groups must be assessed.
  • Whether representative examples and approved data can be accessed during the assessment.
  • The amount of technical investigation or prototype evidence required to resolve uncertainty.

Common questions

Is this an AI strategy workshop?
It is a decision engagement. Discussion provides context, but recommendations are tied to workflows, examples, systems, constraints and named next actions.
Do we need clean data before starting?
No. Data quality and access are part of readiness. We do need enough representative material to identify what is available and what remains unknown.
Will the assessment recommend a custom build?
Only when the evidence supports it. Buying existing software, changing the process or stopping the idea are valid conclusions.
What should we bring first?
Bring one costly or slow workflow, a person who owns it and a few non-sensitive examples that show both routine work and difficult exceptions.

Related evidence

Continue with the underlying work.

Explore AI consulting or See a build-versus-buy decision guide.

Bring us one example

Last updated