A roadmap that lives across seven documents is not a roadmap. It is homework you will not revisit. This compresses the decisions into something that fits on a single page.
Why one page
The value of a plan is not in its completeness — it is in whether you consult it when a decision arises. On a Tuesday evening, when you have twenty minutes and a choice between three tutorials, a forty-page plan is useless. A single page you can scan in fifteen seconds is what actually changes the decision.
Everything in this worksheet has therefore been reduced to the minimum that still guides action. If a field feels too small, that constraint is doing deliberate work.
Fill it in this order
The order matters — each answer narrows the next. Filling it top to bottom takes about forty-five minutes if you have completed the earlier components.
Situation — where you are
Current role, main career goal, current AI stage, biggest challenge, available learning time. This is your starting position, taken from components one and two.
Focus — what you will develop
Three skills and three tasks. Not ten. The number is capped because a list of ten is a wish, and a list of three is a decision.
Method — how you will do it
One main tool and one success metric. If you cannot name the metric, the goal is still too vague to pursue.
Trigger — what happens first
Your first action, small enough to complete today. Not this week. Today.
Be honest rather than optimistic here. Someone who genuinely has fifteen minutes a day and plans for fifteen will outperform someone who claims an hour and delivers twenty minutes twice a week. The plan built on a real number is the one that survives contact with a busy month.
The worksheet
🗺️ My personal AI roadmap
Situation
Beginner · Explorer · Practitioner · Builder
The honest number, not the aspirational one.
Focus
Method
One.
A number you could check.
Trigger
Small enough to finish today.
30 days from today. Calendar it now.
A completed example
Role: Gathers requirements, writes user stories, works between business and engineering.
Career goal: Move into a product ownership role within eighteen months.
AI stage: Explorer (score 21).
Biggest challenge: Uses AI occasionally but nothing has changed in how he works.
Available time: 3 hours per week — 30 minutes on weekdays, occasionally more.
Three skills: structured prompting · requirement gap analysis · summarising stakeholder input.
Three tasks: reviewing requirements for ambiguity · writing acceptance criteria · summarising workshop notes.
Main tool: Claude — chosen for long documents.
Success metric: Requirement clarification requests during sprint reduced from ~8 to under 4 per sprint.
First action: Take last sprint's requirements document and ask what a developer would have to guess.
Note the metric. It is not "get better at AI" — it is a number his team already tracks, that his manager already cares about, and that moves for reasons anyone can trace back to his work.
How to keep it alive
A roadmap decays unless something forces you to look at it. Three mechanisms, in order of effectiveness:
- Physical placement. Print it. Put it where you work. Digital documents in folders are functionally invisible.
- A calendared review. Thirty days out, in your calendar, now. Not a mental note.
- One person who knows. Tell someone your metric. Mild social accountability outperforms private intention by a wide margin.
Change it when evidence says the plan is wrong — the metric proved unmeasurable, the task disappeared, your role shifted. Do not change it because a more exciting option appeared in week two. The distinction between adapting and drifting is whether something you learned forced the change.
Common mistakes
Where this leaves you
You have worked through all eight components. You should now have:
- An honest baseline score and a known weakest dimension
- A role-specific direction rather than generic ambition
- One tool, deliberately chosen, with others deliberately postponed
- A repeatable structure for writing prompts, and the ability to diagnose bad output
- A working prompt library you will extend over time
- Specific use cases mapped to your actual job
- A thirty-day sequence with a deliverable each week
- A one-page roadmap you can act on
That is more structure than most people build in a year of consuming AI content. What determines the outcome now is not more information — it is whether you execute the first action.
Start small. Choose one goal. Use one tool. Improve one real task. Practise consistently.
If you want a roadmap built for your situation
This worksheet gives you a solid general structure. What it cannot do is account for the specifics — your industry, the politics of your team, the constraints of your organisation, or which of your instincts about your own career are accurate.
That is what a one-to-one strategy call is for: working through your actual situation and building the version of this roadmap that fits it.