Start with an outcome
Choose a business problem and define what improvement would look like. Map who does the work, which systems contain the information and what currently causes delay or rework. A list of AI tools is not a strategy.
Prioritise opportunities
Compare candidate projects on potential value, data readiness, integration effort and consequences of failure. Separate training needs from opportunities to build a workflow. Shortlist a pilot with a clear owner, a measurable baseline and a manageable scope.
Plan the dependencies
Identify required permissions, internal reviewers, tool costs and the people who will operate the solution. Set decision points for the pilot, including when to continue, change direction or stop. Match training to the workflow the team will actually use.
A roadmap that supports decisions
The useful output is a sequence of decisions and projects: what comes first, why it matters, what must be ready and how success will be measured. The scope and deliverables are agreed with you; implementation and team training can then be discussed separately.
Frequently asked questions
Should we train people or build first?
If the main barrier is skills, start with training. If a repeatable process is well understood, a small implementation pilot may be appropriate. Often the two work together.