Project objective

This module is the capstone project for the Leadership and Strategy specialization. There is no new theory. Everything you need you have already learned in CX01-CX07. Now you apply it.

The deliverable is an AI strategy document for your company. Not an academic paper. A document you can present to the board of directors, the executive committee or your leadership team that generates concrete decisions.

The document has 10 sections. Each section is explained with its structure, key questions and a template you can fill in. The final result: a 15-25 page document that defines what your company does with AI, why, how, with what resources and with what expected outcomes.

Section 1: Executive summary

The executive summary is written last but read first. It is the only thing 50% of your audience will read. It must be self-contained: if someone only reads this page, they must understand what you propose, why, and what you need.

Structure (1 page, maximum 500 words)

• Context: 2-3 sentences on where the company stands today regarding AI. Maturity level (CX01). Current usage. Uncaptured opportunity.
• Vision: 1-2 sentences on where we want to be in 12 months. Specific, measurable, relevant to the business.
• Key opportunities: the 3 highest-impact opportunities identified. For each: what it does, what impact it has, what it costs.
• Investment required: total budget requested. High-level breakdown: tools, training, talent, infrastructure.
• Expected ROI: estimated return in 12 months. Conservative, not optimistic. Better to surprise upward than to disappoint.
• Decision requested: what you need the board to approve. Budget, responsible party, deadlines. Explicit.

Section 2: Maturity assessment

Use the 5-level framework from CX01. Be honest. Overestimating your maturity creates unrealistic expectations and projects that fail.

What to assess

The overall level is the average of the 6 dimensions, rounded down. If you have a 4 in usage and a 1 in governance, your real level is not 2.5. It is 1, because governance holds back everything else.

Critical gaps

Identify the 2-3 dimensions with the lowest scores. Those are your critical gaps. The action plan must prioritize them. There is no point investing in advanced tools (dimension 5) if you lack governance (dimension 2).

Section 3: Opportunity map

List all AI opportunities you have identified throughout this specialization. For each one, document:

Opportunity brief

• Name: descriptive, no jargon. "Automate monthly sales reports", not "Implement NLG pipeline".
• Department: which benefits directly.
• Current problem: what pain it solves. How much time/money it costs today.
• AI solution: how AI solves it. Which tool or model.
• Estimated impact: in hours saved, cost reduced or revenue generated. Be conservative.
• Effort: what you need to implement it. Tools, training, integration, time.
• Horizon: H1 (0-6 months), H2 (6-18 months), H3 (12-36 months).
• Dependencies: do you need something first? Clean data, trained team, approved tool.

Aim to identify 10-15 opportunities. Not all will be executed. But you need critical mass to prioritize well.

Section 4: Prioritization (impact vs effort)

With 10-15 opportunities identified, you need to decide which to execute first. Use the impact vs effort matrix:

How to score

Impact (1-10): how much value does this opportunity generate? Consider time savings, cost reduction, new revenue, quality improvement, competitive advantage.

Effort (1-10): how much does it cost to implement? Consider time, money, technical complexity, organizational change, dependencies.

The 4 quadrants

• High impact, low effort (quick wins): execute immediately. These are your first victories. They generate confidence and budget for larger projects.
• High impact, high effort (strategic projects): plan and execute in H2-H3. They need resources, but the return justifies them.
• Low impact, low effort (filler): execute if resources are left over, but never at the expense of quick wins or strategic projects.
• Low impact, high effort (discarded): do not execute. They consume resources without proportional returns.

Select 3-5 quick wins for the first 90 days. 2-3 strategic projects for the year. Discard the rest without guilt.

Section 5: 12-month roadmap

The roadmap translates prioritization into a calendar. It is not a Gantt chart with 200 tasks. It is a high-level plan with clear milestones.

Quarterly structure

Q1 (months 1-3): Foundations

• Approve AI policy and designate responsible party
• Complete AI systems inventory and AI Act classification
• Basic training (Level 1) for the entire team
• Execute 3 identified quick wins
• Measure baseline for key metrics

Q1 milestone: policy approved, team trained, 3 processes improved with AI, baseline metrics established.

Q2 (months 4-6): Scale

• Advanced training (Level 2) for champions and managers
• Implement AI productivity tools across all departments
• Launch the first strategic project (H2)
• Evaluate AI Act compliance for high-risk systems
• First metrics review vs baseline

Q2 milestone: adoption >40%, first strategic project underway, compliance evaluated.

Q3 (months 7-9): Deepen

• Evaluate data sovereignty and cloud/self-hosted decisions
• Complete first strategic project, launch second
• Pilot new AI product/service (if applicable, H2-H3)
• Adoption audit and plan adjustment

Q3 milestone: adoption >60%, first strategic project completed, pilot launched.

Q4 (months 10-12): Consolidate

• Evaluate year-end results: actual ROI vs estimated
• Scale projects that work, kill those that do not
• Prepare Year 2 AI strategy (based on real learnings)
• Present results to the board

Q4 milestone: ROI demonstrated, Year 2 plan prepared, board presentation delivered.

Section 6: Budget and allocation

The budget must be specific, justified and realistic. Board members do not approve "investing in AI". They approve "EUR 120,000 for training, tools and infrastructure that will generate an estimated EUR 300,000 in savings over 12 months".

Spending categories

How to justify the investment

For every euro requested, show the expected return. Not in generic terms ("improve productivity") but in concrete numbers:

• "Training 100 employees (EUR 30K) will save 2 hours/person/week. At an average cost of EUR 25/hour, that is EUR 260K/year in freed productivity."
• "Automating tier-1 support (EUR 15K for the tool) will resolve 40% of tickets without human intervention. With 500 tickets/month, that saves 200 agent hours/month."
• "The AI product pilot (EUR 20K) has the potential to generate EUR 150K in revenue over 12 months if we validate with 5 pilot clients."

Section 7: Governance structure

Governance defines who decides what about AI. Without governance, each department does what it wants, buys tools without coordination and generates risks nobody controls.

Minimum roles

• Executive sponsor: a member of the executive committee who backs the AI strategy. Does not need technical expertise. Needs authority and commitment.
• AI lead: the person who executes the strategy day to day. Can be a CAIO, a Head of Innovation, a CTO with an expanded mandate. Needs dedicated time (minimum 50%).
• AI committee: monthly meeting with representatives from Legal, IT, HR, Business and the AI lead. Decides: which tools are approved, which projects launch, which risks are escalated.
• Department champions: 1 per 20-30 employees (CX04). They are the committee's execution arm in each area.

Minimum policies

1. Acceptable use policy: which tools are approved, what data can be shared, what is not permitted.
2. Acquisition policy: process for evaluating and approving new AI tools. Who requests, who evaluates, who approves.
3. Data policy: data classification for AI use (public, internal, confidential, prohibited). Aligned with GDPR and the AI Act.
4. Incident policy: what to do when AI makes an error with impact. Who is notified, how it is escalated, how it is documented.

Section 8: KPIs and metrics

Define between 5 and 10 KPIs that demonstrate the AI strategy's progress. Each KPI must have: definition, calculation formula, baseline, 12-month target and measurement frequency.

Recommended KPIs

Section 9: Risk assessment

Every plan has risks. Documenting them is not pessimism, it is professionalism. The board wants to know you have thought about what can go wrong and that you have a plan for it.

Risk framework

For each risk: who monitors it, how often it is reviewed, what trigger activates the mitigation plan.

Section 10: Board presentation

The strategy document is the complete reference. The board presentation is the pitch. They are not the same. The presentation is what persuades. The document is what gets consulted afterward.

Presentation structure (10-15 slides, 30 minutes)

1. Slide 1: Title and context (30 sec). "AI Strategy [Company] 2026-2027".
2. Slide 2: Why now (2 min). Three urgent reasons: competition, regulation, market opportunity.
3. Slide 3: Where we stand (2 min). Maturity level. Current usage. Critical gaps.
4. Slide 4: Where we are heading (2 min). 12-month vision. Expected outcome. KPIs.
5. Slides 5-7: The 3 key opportunities (6 min). One slide per opportunity: problem, solution, impact, cost.
6. Slide 8: Roadmap (2 min). 4 quarters with key milestones. Visual, not text.
7. Slide 9: Budget (2 min). Investment required vs expected return. One clear table.
8. Slide 10: Risks and mitigation (2 min). Top 3 risks with a plan. Shows you have thought about what can go wrong.
9. Slide 11: Governance (1 min). Who leads, how decisions are made, when reviews happen.
10. Slide 12: Decision requested (2 min). What you need them to approve. Budget, responsible party, deadlines. Explicit.

Rules for the presentation

• Less text, more numbers. Board members do not read paragraphs on slides. They want figures, charts, comparisons.
• Talk business, not technology. Do not mention model names or architectures. Talk about outcomes: savings, revenue, risk, competitiveness.
• Anticipate questions. The 3 questions they always ask: "how much does it cost?", "how much do we save?" and "what happens if we do nothing?". Have the answers ready with numbers.
• Ask for something concrete. "We request approval of EUR 120K in budget, the designation of [name] as AI lead, and a quarterly progress review at the executive committee."

Final checklist

Puntos clave