Chapter 39 Exercises: Capstone — AI Transformation Plan

The Capstone Project

The following exercises form the capstone project itself. Exercises 39.1 through 39.10 correspond to the ten components of the AI Transformation Plan described in Section 39.1. Complete all ten in sequence to produce a board-ready strategic deliverable.

Exercises 39.11 through 39.15 are supplementary component exercises that deepen your analysis of individual elements. Your instructor may assign these as standalone exercises or as extensions to the capstone.

Section A: The Capstone Project

Exercise 39.1: Industry Selection and Analysis Select an organization for your AI Transformation Plan (see Section 39.2 for options). Produce a 2-3 page AI Landscape Assessment for your chosen industry and organization that includes: - (a) Industry AI adoption maturity — where does this industry sit on the adoption curve? Cite at least two published industry surveys. - (b) Competitive landscape — identify 3-5 competitors and their publicly documented AI initiatives. Assess whether the competitive gap is widening or narrowing. - (c) Regulatory environment — identify the 3-5 most relevant regulations (industry-specific and AI-specific) that constrain or shape AI adoption. - (d) Data ecosystem — describe the major data assets the organization generates, purchases, or has access to. Assess the likely quality, accessibility, and governance maturity of these data assets. - (e) Workforce readiness — estimate the current level of AI literacy, existing technical talent, and predominant workforce attitude toward AI. - (f) Top three AI opportunities and top three barriers, with justification.

Exercise 39.2: AI Maturity Assessment Using the AIMaturityAssessment tool (or a manual equivalent), conduct a maturity assessment for your chosen organization: - (a) Score the organization on all six dimensions (strategy, data, technology, talent, governance, culture) on a 1-5 scale. Provide a one-paragraph justification for each score. - (b) Set target scores for a 24-month horizon. Justify each target — why is this the right level of ambition? - (c) Run the gap analysis. Identify the three dimensions with the largest gaps. For each, explain why the gap exists and what structural barriers must be overcome to close it. - (d) Run the benchmark comparison against industry averages. Where does your organization sit relative to peers? What does this imply for the urgency and ambition of the transformation plan? - (e) Run the code and include the output in your submission.

Exercise 39.3: Use Case Identification Using the AI Opportunity Canvas (Section 39.4), brainstorm at least ten AI use cases for your chosen organization: - (a) Complete all seven fields of the Opportunity Canvas for each use case. - (b) Ensure your use cases span at least three categories: customer-facing, operational, back-office, and strategic/analytical. - (c) For each use case, identify the primary AI/ML technique required (classification, regression, NLP, computer vision, recommendation, forecasting, generative AI, etc.) and reference the relevant chapter in this textbook. - (d) For at least two use cases, describe the ethical considerations in detail — what could go wrong, and who would be harmed?

Exercise 39.4: Use Case Prioritization Using the Impact-Feasibility Matrix (Section 39.5), prioritize your use cases: - (a) Score each use case on Impact (1-10) and Feasibility (1-10) using the sub-criteria defined in the chapter. Show your scoring breakdown. - (b) Plot all use cases on the Impact-Feasibility Matrix and classify each into a quadrant (Quick Win, Strategic Bet, Fill-In, Deprioritize). - (c) Construct a portfolio: select 2-3 Quick Wins for Phase 1, 2-3 Foundation items for Phase 2, 2-3 Scale items for Phase 3, and 1-2 Optimize items for Phase 4. Justify your selections. - (d) Identify at least one use case you chose to not pursue despite high impact. Explain why. - (e) For each Phase 1 use case, write a two-paragraph business case that a CFO would find compelling.

Exercise 39.5: Technology Architecture Design Design the technology architecture for your AI transformation: - (a) For each technology layer (cloud platform, data platform, ML platform, GenAI platform, MLOps, monitoring), recommend a specific solution. Justify each selection using the build-buy-configure framework from Chapter 22. - (b) Draw (or describe in structured text) an architecture diagram showing how the components interact. - (c) Address the five data architecture principles from Section 39.6: single source of truth, quality by design, access with governance, privacy by design, and scalability. - (d) Estimate the Year 1 technology cost across the architecture layers. Break down costs by license/subscription, implementation, and ongoing operations. - (e) Identify the top two vendor lock-in risks and propose mitigation strategies.

Exercise 39.6: Governance Framework Design Design an AI governance framework for your chosen organization: - (a) Define three risk tiers with specific criteria for classifying use cases into each tier. Provide at least two example use cases for each tier from your portfolio. - (b) For each tier, specify the governance requirements: who reviews, what documentation is required, what testing must be completed, and what monitoring is mandated. - (c) Design an AI ethics review process: who sits on the review board, how often does it meet, what triggers a review, and what authority does it have? Reference the frameworks from Chapter 27. - (d) If your industry has specific regulatory requirements for AI (e.g., FDA for healthcare, SR 11-7 for financial services, EU AI Act high-risk classification), explain how your governance framework addresses them. - (e) Address the shadow AI challenge (Chapter 22): how will you manage unauthorized AI tool usage? Design a citizen data science program with appropriate guardrails.

Exercise 39.7: Implementation Roadmap Create a phased implementation roadmap: - (a) Using the TransformationRoadmapGenerator (or a manual equivalent), assign use cases to phases and generate a timeline. Include the output in your submission. - (b) Map dependencies between initiatives. Identify at least three dependency chains where one initiative must complete (or reach a milestone) before another can begin. - (c) For each phase, specify: key milestones, resource requirements (team size, skills needed), infrastructure deliverables, and governance checkpoints. - (d) Define stage-gate criteria: what must be true for the organization to advance from Phase 1 to Phase 2? From Phase 2 to Phase 3? What would trigger a pause or pivot? - (e) Identify the critical path — the sequence of activities that determines the minimum total timeline. What happens if an activity on the critical path is delayed?

Exercise 39.8: Change Management Plan Develop a change management plan: - (a) Complete a stakeholder analysis using the Power-Interest Grid (Chapter 35). Identify at least eight stakeholder groups and place them on the grid. - (b) For each stakeholder group, design a differentiated engagement strategy: communication channels, frequency, messaging, and success metrics. - (c) Design a three-level training program (AI Awareness, AI Literacy, AI Proficiency). For each level, specify: target audience, duration, content topics, delivery format, assessment method, and certification (if applicable). - (d) Identify the three most likely sources of resistance in your organization. For each, develop a specific mitigation strategy that goes beyond generic advice. Reference the Prosci ADKAR model from Chapter 35. - (e) Design a communications plan for the first 90 days of the transformation. Include: launch announcement, stakeholder briefings, town halls, and internal marketing/branding for the AI initiative.

Exercise 39.9: Financial Analysis Produce a financial analysis for the AI transformation: - (a) Estimate the total investment across five cost categories (technology, talent, data, change management, governance) for each of the four phases (24-month horizon). Present as a phased budget table. - (b) Project the annual value for each use case in the portfolio. Distinguish between direct value (revenue, cost savings, risk reduction) and indirect value (productivity, experience, competitive positioning). - (c) Calculate the overall ROI for the transformation at Year 1, Year 2, and Year 3 (assuming ongoing value beyond the 24-month implementation period). - (d) Present three scenarios: optimistic, base, and pessimistic. For each scenario, state the key assumptions and their justification. - (e) Apply risk-adjusted analysis: discount the base scenario for probability of failure and implementation delay using the formula from Section 39.10. State the discount factors you chose and justify them. - (f) Calculate the payback period — how many months until cumulative value exceeds cumulative investment?

Exercise 39.10: Executive Summary and Presentation Produce the final capstone deliverables: - (a) Write a one-page executive summary following the structure in Section 39.15 (opportunity, current state, plan, return, ask, risk). This should be a standalone document that could be handed to a board member. - (b) Create a 10-slide presentation deck. Suggested structure: (1) Title, (2) The Opportunity, (3) Current State (Maturity Assessment), (4-5) Use Case Portfolio, (6) Technology Architecture, (7) Governance Framework, (8) Roadmap and Financials, (9) Risks and Mitigation, (10) The Ask. - (c) Prepare for a 15-minute presentation with 10 minutes of Q&A. Anticipate the three most likely questions an evaluator would ask and prepare concise answers. - (d) Include a one-paragraph "what I don't know" section that honestly identifies the gaps in your analysis and how you would address them with more time or information.

Section B: Component Exercises

Exercise 39.11: Cross-Textbook Synthesis For each of the following chapters, explain how its content directly informs one specific component of your capstone plan. Be specific — reference a particular framework, tool, or concept from the chapter and explain how you applied it: - (a) Chapter 4 (Data Strategy and Data Literacy) - (b) Chapter 11 (Model Evaluation and Selection) - (c) Chapter 22 (No-Code / Low-Code AI) - (d) Chapter 25 (Bias in AI Systems) - (e) Chapter 27 (AI Governance Frameworks) - (f) Chapter 31 (AI Strategy for the C-Suite) - (g) Chapter 34 (Measuring AI ROI) - (h) Chapter 35 (Change Management for AI)

Exercise 39.12: Maturity Assessment Comparison Using the AIMaturityAssessment tool, assess three organizations in different industries: - (a) Run the assessment for each organization. Compare the radar charts. - (b) Identify which dimensions show the most variation across industries and which are consistently low. What does this pattern suggest about the state of enterprise AI? - (c) For the organization with the lowest overall maturity, develop a "90-day quick start" plan that addresses the most critical gap. - (d) For the organization with the highest overall maturity, identify the plateau risks — where might progress stall, and why?

Exercise 39.13: Competitive Response Scenario Your organization has completed Phase 1 of its AI transformation when a major competitor announces a partnership with a leading AI company (similar to the NovaMart scenario from Athena's experience). The board is pressuring the CEO to accelerate the AI roadmap and skip Phase 2 to move directly to Phase 3. - (a) What are the risks of accelerating and skipping Phase 2 (Foundation)? - (b) What would you recommend to the CEO? How would you balance the urgency of competitive pressure with the discipline of a phased approach? - (c) Drawing on Ravi's retrospective lesson #5 (competitive pressure as a catalyst for disciplined adoption), draft a one-page memo to the board arguing for an accelerated but not reckless response.

Exercise 39.14: Ethical Dilemma — The Tempting Use Case During Phase 2, your AI team identifies a use case that was not in the original plan: using employee email and calendar data to predict which employees are likely to leave (attrition prediction). The VP of HR is enthusiastic — voluntary attrition costs the organization millions annually. The use case has high impact and moderate feasibility. - (a) Apply the governance framework you designed in Exercise 39.6 to this use case. What risk tier does it fall into? - (b) What ethical concerns does this use case raise? Consider privacy, consent, power dynamics, and potential for misuse. Reference frameworks from Chapter 25 (bias), Chapter 26 (explainability), and Chapter 29 (privacy). - (c) Should the organization proceed with this use case? If yes, what safeguards are necessary? If no, what alternative approach could address the attrition problem without the ethical concerns? - (d) Draft a one-paragraph recommendation for the AI ethics review board.

Exercise 39.15: Post-Mortem Simulation It is 18 months after your AI transformation began. Phase 1 succeeded — the quick wins delivered 70% of projected value. But Phase 2 is behind schedule because data integration took twice as long as expected, the governance framework was not adopted consistently across business units, and the AI team has experienced 40% turnover. - (a) Conduct a retrospective analysis. For each of the three problems (data integration delays, governance adoption, talent retention), identify the root cause and propose a corrective action. - (b) Should the organization proceed to Phase 3 as planned, delay Phase 3 until Phase 2 is complete, or modify the Phase 3 scope? Justify your recommendation. - (c) Write a candid one-page update for the board that communicates both the Phase 1 success and the Phase 2 challenges. How do you maintain board confidence while being honest about setbacks? - (d) Reflecting on this scenario, what would you add to your original plan (Exercise 39.7) to reduce the probability of these problems occurring?

The capstone project (Exercises 39.1 through 39.10) is designed to take approximately 30-50 hours. It may be completed individually or in teams of two to three. If completed in teams, each team member should contribute to all ten components, not divide and conquer — the integrative thinking is the point. Refer to Chapter 40 for the closing reflection on what kind of AI leader you intend to become.