AI Project Planner | Leila Fielding

Runs in your browser

Use this when: scoping an AI project across delivery, roles, evidence, tasks and review gates—not just testing a single prompt.

WHAT YOU'LL ACHIEVE

Shape the problem, scope, evidence, responsibilities, milestones and human decision points before work begins.

THE TOOL: PROJECT PLAN

Start with the problem, not the model.

Guided planning only. Costs are your estimates; this tool does not price, execute, approve or decide whether to launch.

Project objective / problemWhat problem should this project solve?Users and stakeholdersWho is affected, who uses the outcome, and who can raise concerns?Current processHow is this work done today? Include volumes, handoffs and pain points.Desired outcomesWhat should improve for users or the organisation?Measurable success metricsName a baseline, target and measurement method where possible.AI suitabilityWhy might AI help here compared with a non-AI change? What evidence would support that?Scope boundariesWhat must remain outside the AI's role? What should it never decide or do?Data availability and qualityWhat data exists, who owns it, and what quality gaps are known?Sensitivity and permissionsCould data be personal, confidential or sensitive? What permissions and approved environments apply?Dependencies and integrationsWhich systems, suppliers, people, approvals or interfaces are needed?Budget assumptions (your estimates)Enter only your own assumptions; no live prices are retrieved.Cost range / cost ownerYour estimate, currency, basis and who will validate it.Evaluation criteriaHow will quality, safety, usefulness and failure handling be evaluated?Change management and trainingWho needs to be informed, consulted, trained or supported?Governance and human oversightWho is accountable? When can a person review, pause, correct or reverse an outcome?Review date and decision checklistSet a review date and list evidence required before a human go/no-go discussion.

PEOPLE AND ACCOUNTABILITY

Name owners before tasks begin.

RACI-style roles are prompts, not a substitute for your organisation’s governance.

People responsible, accountable, consulted and informed
Role	Responsible	Accountable	Consulted	Informed	Remove

HUMAN DECISION CHECKLIST

Keep unresolved decisions visible.

These statuses are supplied by you. “Ready” is not an automated approval.

Scope and boundariesData permissionsRisk reviewEvaluation evidenceHuman oversight

Planning fields still need input: Project objective, Affected users / stakeholders, Current process, Desired outcomes, Success measures, AI scope boundaries, Data availability, Data sensitivity and permissions, Accountable human oversight, Evaluation criteria, Review date / decision evidence.

PHASED MILESTONES AND TASKS

Build an ordered, editable plan.

Dependencies must point to another current task. Reorder is blocked if it would place a dependency after its dependent.

No tasks yet. Add the suggested phases or create a task to start your timeline.

REVIEW AND EXPORT

Review gaps with the people who own them.

For evaluation, see the Testing Checklist.

AI PROJECT PLAN

Project objective / problem
Not yet specified.

Users and stakeholders
Not yet specified.

Current process
Not yet specified.

Desired outcomes
Not yet specified.

Measurable success metrics
Not yet specified.

AI suitability
Not yet specified.

Scope boundaries
Not yet specified.

Data availability and quality
Not yet specified.

Sensitivity and permissions
Not yet specified.

Dependencies and integrations
Not yet specified.

Budget assumptions (your estimates)
Not yet specified.

Cost range / cost owner
Not yet specified.

Evaluation criteria
Not yet specified.

Change management and training
Not yet specified.

Governance and human oversight
Not yet specified.

Review date and decision checklist
Not yet specified.

PEOPLE / RACI-STYLE OWNERSHIP
Sponsor: Responsible ; Accountable ; Consulted ; Informed
Project owner: Responsible ; Accountable ; Consulted ; Informed
Data steward: Responsible ; Accountable ; Consulted ; Informed
Risk / privacy reviewer: Responsible ; Accountable ; Consulted ; Informed
Evaluation lead: Responsible ; Accountable ; Consulted ; Informed
Change lead: Responsible ; Accountable ; Consulted ; Informed

DECISION CHECKLIST
scope: Unknown
data: Unknown
risk: Unknown
evaluation: Unknown
oversight: Unknown

TASKS / MILESTONES
No tasks added yet.

OPEN ITEMS
- Project objective
- Affected users / stakeholders
- Current process
- Desired outcomes
- Success measures
- AI scope boundaries
- Data availability
- Data sensitivity and permissions
- Accountable human oversight
- Evaluation criteria
- Review date / decision evidence

This is a planning brief, not legal advice, a compliance finding, risk clearance or an automated go-live decision. A human decision-maker must review evidence, risks, costs and approvals. Estimates are user supplied. Inputs remain in this browser session and are not saved by this tool.