About the AI Project Scoper

Who it is for

Use this tool when the decision to build is made and you need a first plan. It is for the person who has to turn an approved idea into phases, tasks and roles: a programme lead, an engineering manager or a head of AI.

Set the context first: the AI approach, your domain, team size, timeline and budget. Then answer the scoping questions. Allow fifteen to twenty minutes.

What it measures

The scoping questions cover six dimensions, and the plan is tailored to the approach and context you selected.

Data and infrastructure. Sources, quality, pipelines and the platform they run on.
Model and approach fit. Whether the chosen approach suits the problem, and what it takes to prototype it.
Architecture and integration. The systems the AI component connects to and the non-functional requirements.
Governance, ethics and risk. Regulation, privacy, fairness and risk ownership.
Team and change management. Skills, roles and the effect on existing ways of working.
Evaluation and testing. Success metrics, baselines and how quality is verified.

How to read the result

The result is a roadmap in five phases: discovery, prototype, validate, deploy and scale. Each phase lists the tasks that apply to your answers, with a suggested role and a size.

Tasks in your weakest dimensions are raised in priority, so the work that reduces the most risk comes first. The risk view groups the main risks by dimension.

Export the plan as an Excel workbook with a summary, the roadmap, the task list and the risks. Treat it as a first draft to adapt with your team, not a commitment.

Common questions

How do you scope an AI project?

Fix the approach and the constraints first, then assess each area that can block delivery: data, model fit, architecture, governance, team and evaluation. The plan follows from the gaps. Scoping from the technology outward is how projects end up with a working demo and no path to production.

What are the phases of an AI project?

Discovery to frame the problem and audit the data, a prototype to prove the approach, validation against real cases and success metrics, deployment into production, and scaling with monitoring and governance in place.

I have not chosen an AI approach yet. Where do I start?

Run the AI Approach Finder first. It recommends an approach in a few minutes, and you can bring that choice into the scoper.

Is it free?

Yes. The tool is free and runs in your browser with no account. The downloadable report asks for an email address, and your scores are sent with it so I can follow up if you want to talk them through.