AI readiness assessment: what to check before you invest

An AI readiness assessment examines whether a proposed AI project has a clear business purpose, suitable data and a practical route into everyday use. It should help you decide whether to proceed, run a limited trial or address other issues first.

If your business is considering AI, start with a specific task you want to improve. That gives you something concrete to assess before committing to software, infrastructure or development.

Identify the task and the problem

Give the proposed project a clear starting point.

For example, a team might spend time finding answers across internal documents, transferring information between systems or preparing routine summaries.

For each task, establish:

• Who carries it out and how often.
• How much time it currently takes.
• Where delays, errors or rework occur.
• What a successful outcome would look like.

Be precise about the cause of the problem. If staff cannot find information because documents are outdated or contradictory, an AI tool will still need a reliable source of information to work from.

Consider whether AI is the right approach

Compare AI with the other practical options.

A standard integration might remove duplicate data entry. A clearer form might reduce incomplete submissions. A change to an existing system might eliminate an unnecessary approval step.

AI is worth investigating where the task involves interpreting varied information, working with language or producing a draft that a person can review. Even then, the useful solution may combine AI with conventional software and human judgement.

The assessment should leave room for a straightforward recommendation: solve this particular problem without AI.

Check the information the system would need

List the documents, records and systems required for the proposed task.

Then consider whether the information is current, sufficiently complete and accessible to the people who need it. Identify differences in access permissions: a tool that searches company information must respect who is entitled to see each source.

Questions to resolve include:

• Where does the information live?
• Who owns and maintains it?
• Which users should be able to access it?
• What information should be excluded?
• How will updates and deletions reach the new system?

These answers help define the project’s scope and the work needed before a trial.

Plan how it would fit into everyday work

A useful demonstration is only one part of implementation.

Consider where staff would use the system, how it would connect to existing tools and what should happen when an answer is incomplete or a task fails.

For an internal document assistant, for example, you might require links to the source material, clear handling of unanswered questions and a route for staff to flag inaccurate responses.

Decide who will review outputs and who will be responsible for maintaining the system after launch.

Compare deployment options against your requirements

Hosting should follow the requirements of the task.

Document the controls you need over data access, processing, storage, support and system changes. Use those requirements to compare suitable managed services and private deployments.

Include the operational responsibilities in that comparison. Someone needs to manage access, monitor performance, maintain integrations and assess changes to the underlying model or service.

Strategies provides guidance on AI strategy and integration, alongside private AI hosting where that fits the requirement.

Define a trial with measurable success criteria

Choose one workflow and a manageable group of users.

Measure the existing process before introducing AI, then compare results using representative tasks. Include difficult cases as well as straightforward examples.

Useful measures could include:

• Time needed to complete the task, including checking and correction.
• The proportion of outputs that meet agreed quality criteria.
• How often a person needs to intervene.
• Whether staff use the system in their normal work.
• The ongoing cost of delivering the result.

Agree in advance what would justify expanding the trial, changing the approach or stopping.

Create a short brief before commissioning work

You can begin with six points:

1. The task you want to improve.
2. The people affected by the problem.
3. The systems and information involved.
4. The current time, cost or quality issue.
5. The constraints the solution must work within.
6. The evidence you would need to approve further investment.

Any unanswered points become questions for discovery.

Discuss your proposed AI project

Strategies helps businesses assess AI opportunities and plan development, integration and deployment.

If you have a process you would like to improve, contact our team with a short description of how it works today and what you want to change. We can discuss the requirements and an appropriate next step.