Chapter 33 Exercises: Project Planning and Estimation

These exercises are organized into five tiers based on Bloom's taxonomy, progressing from basic recall to creative challenges.

Tier 1: Recall (Exercises 1-6)

These exercises test your ability to remember key concepts from the chapter.

Exercise 1: The Asymmetric Acceleration Principle

Define the Asymmetric Acceleration Principle in your own words. List the five phases of feature development and their typical AI acceleration factors as presented in Section 33.1.

Exercise 2: Three Acceleration Tiers

Name the three acceleration tiers used for task classification. For each tier, provide the acceleration factor range and at least two example task types.

Exercise 3: AI-Specific Risks

List the five AI-specific project risks discussed in Section 33.6. For each risk, write one sentence describing why it matters.

Exercise 4: Sprint Ceremony Adaptations

For each of the four Scrum ceremonies (Sprint Planning, Daily Standup, Sprint Review, Sprint Retrospective), list at least one adaptation recommended for AI-augmented teams.

Exercise 5: Key Estimation Terms

Define the following terms as they are used in this chapter: - AI-atomic task - Three-Point AI Estimation Method - AI Sprint Coefficient - Quality-Adjusted Velocity (QAV) - First-Prompt Success Rate

Exercise 6: Stakeholder Communication Scenarios

Describe the three common stakeholder communication scenarios presented in Section 33.9 and summarize the recommended response framework for each.

Tier 2: Comprehension (Exercises 7-12)

These exercises test your understanding of the concepts and their relationships.

Exercise 7: Explain the Planning Paradox

In Section 33.1, the chapter introduces the Planning Paradox. Explain in your own words why planning becomes more important in the AI era, not less. Use a specific numerical example to illustrate your point (for example, a project where implementation previously consumed 60% of total effort).

Exercise 8: The Velocity Trap

Explain the Velocity Trap described in Section 33.1. Why is it dangerous for a team to set expectations based on their peak AI-assisted productivity? What are the downstream consequences, and how should a team respond?

Exercise 9: Burndown Pattern Analysis

The chapter describes a "hockey stick" burndown pattern that is common in AI-augmented sprints. Explain why this pattern occurs and describe three strategies for addressing stakeholder concerns about it.

Exercise 10: Code Review Bottleneck

Explain why code review can become a bottleneck in AI-augmented projects even though the overall team is producing more code. What specific metrics would you track to detect this bottleneck early? Propose two solutions.

Exercise 11: Phase Distribution Shift

The chapter shows how AI changes the relative distribution of effort across Waterfall phases. Explain why requirements and design phases become proportionally larger even though their absolute duration may not change. How does this shift relate to long-standing recommendations from software engineering literature?

Exercise 12: Portfolio Effect

Explain the Portfolio Effect described in Section 33.4. Why do AI estimation errors tend to cancel out across many tasks even though individual task estimates have higher variance? Under what conditions might the Portfolio Effect fail to smooth out errors?

Tier 3: Application (Exercises 13-20)

These exercises ask you to apply the concepts to practical situations.

Exercise 13: Task Classification

Classify the following tasks into the three AI acceleration tiers. Justify each classification:

a) Building a REST API for user profile management (CRUD operations) b) Designing the database schema for a multi-tenant SaaS application c) Writing unit tests for an existing payment processing module d) Conducting user interviews to understand workflow requirements e) Implementing a custom sorting algorithm optimized for a specific data distribution f) Creating deployment scripts for a Kubernetes cluster g) Writing API documentation from existing code h) Reviewing and approving a pull request for security-sensitive code i) Scaffolding a React component library with Storybook j) Negotiating API contracts with a third-party vendor

Exercise 14: Three-Point Estimation

Using the Three-Point AI Estimation Method, estimate the following tasks. Provide optimistic, realistic, and pessimistic estimates for both traditional and AI-augmented approaches, then calculate the PERT estimate for each.

a) Building a user registration and login system with email verification b) Creating a dashboard that displays real-time analytics charts c) Implementing a file upload system with virus scanning and cloud storage d) Writing comprehensive integration tests for a REST API with 15 endpoints

Exercise 15: Sprint Capacity Calculation

A team of four developers is planning a two-week sprint. Each developer is available for 6 productive hours per day. The sprint backlog contains: - 5 Tier 1 tasks (estimated at 8 hours each without AI) - 3 Tier 2 tasks (estimated at 12 hours each without AI) - 2 Tier 3 tasks (estimated at 16 hours each without AI)

Using acceleration factors of 4x for Tier 1, 2x for Tier 2, and 1.1x for Tier 3, calculate: a) Total effort without AI assistance b) Total effort with AI assistance c) Whether the sprint is feasible d) How much spare capacity exists (if any) and what you would recommend doing with it

Exercise 16: Risk Assessment Matrix

You are leading a project to build an e-commerce platform. The team has been using AI coding tools for two months. Complete a risk assessment matrix with at least eight risks (including both traditional and AI-specific risks). For each risk, assign probability (1-5), impact (1-5), calculate the risk score, and propose a specific mitigation strategy.

Exercise 17: MoSCoW Reprioritization

A product team has the following feature backlog with traditional effort estimates:

Apply AI acceleration factors (Tier 1: 4x, Tier 2: 2x, Tier 3: 1.2x) to recalculate effort. Then perform a MoSCoW categorization for a 12-week project timeline. Explain your reasoning.

Exercise 18: Stakeholder Presentation

Write a one-page memo to a non-technical executive explaining why your team's project timeline is 35% shorter than similar past projects. Use the Three Timelines approach and address potential skepticism. The project is a customer portal with user management, order tracking, and a support ticket system.

Exercise 19: Burndown Chart Construction

Given the following sprint data, construct a burndown chart (you may describe it textually or as a table). Identify the point where AI-accelerated tasks are exhausted and non-acceleratable tasks dominate.

Exercise 20: Coding Exercise -- Task Decomposer

Using the TaskDecomposer class from code/example-01-task-decomposer.py as a reference, write a function that takes a list of tasks (each with name, description, and estimated hours) and returns a categorized breakdown with AI acceleration estimates. Your function should:

a) Classify each task into an AI acceleration tier based on keywords in the description b) Calculate the AI-adjusted estimate for each task c) Return a summary showing total original hours, total adjusted hours, and overall acceleration factor

Tier 4: Analysis and Evaluation (Exercises 21-26)

These exercises require critical thinking and judgment.

Exercise 21: Methodology Comparison

Compare how AI acceleration affects Scrum versus Kanban workflows. For each methodology: a) Identify two specific advantages of the methodology in an AI-augmented context b) Identify two specific disadvantages or challenges c) Recommend which type of project or team would benefit most from each methodology when AI tools are in use

Exercise 22: Estimation Accuracy Analysis

A team has completed five sprints with the following data:

Analyze this data: a) What is the team's average velocity and estimation accuracy? b) Is there a correlation between AI-accelerated percentage and defect count? c) What would you recommend for Sprint 6 capacity planning? d) Calculate the Quality-Adjusted Velocity for each sprint (assume each defect costs 2 story points to resolve).

Exercise 23: Risk Cascade Analysis

Imagine a scenario where an AI tool provider announces a 3x price increase effective in 60 days, midway through a 6-month project. Analyze the cascading effects of this event on: a) Budget b) Timeline c) Team morale and productivity d) Stakeholder confidence e) Technical decisions

Propose a mitigation plan with specific actions and timelines.

Exercise 24: The Overcounting Problem

A project manager reports that the team's velocity has increased from 40 points per sprint (pre-AI) to 120 points per sprint (with AI). The team size has not changed. Critically evaluate this claim: a) What might explain such a dramatic increase? b) What are three possible measurement errors or misinterpretations? c) How would you validate whether the reported velocity increase reflects genuine productivity improvement? d) What risks does a 3x velocity claim create for future planning?

Exercise 25: Cross-Team Dependency Analysis

In a SAFe environment, three teams are working on a shared product. Team A (building APIs) uses AI heavily and has seen 3x acceleration. Team B (building the frontend) uses AI moderately and has seen 2x acceleration. Team C (building the data pipeline) uses AI minimally and has seen only 1.3x acceleration. Analyze: a) How do the different acceleration rates affect cross-team dependencies? b) What problems might arise during PI Planning? c) How should the Release Train Engineer adapt coordination strategies? d) What metrics should the ART track to ensure alignment?

Exercise 26: Ethical Evaluation

A startup CEO proposes the following: "Since AI lets us build software 5x faster, we should lay off 60% of our development team and keep only the best prompt engineers." Critically evaluate this proposal from the perspectives of: a) Technical feasibility b) Business risk c) Team dynamics and knowledge retention d) Long-term product quality e) Ethical considerations

Tier 5: Creation (Exercises 27-30)

These exercises challenge you to create original artifacts.

Exercise 27: AI-Augmented Planning Playbook

Create a comprehensive planning playbook for a team of six developers who are about to adopt AI coding tools for the first time. Your playbook should include: a) A 4-week transition plan b) Metrics to track during the transition c) Template for AI-augmented sprint planning d) Guidelines for when to use and when not to use AI e) An escalation process for when AI-generated code causes problems

Exercise 28: Estimation Calibration Tool

Design (in pseudocode or Python) an estimation calibration tool that: a) Accepts historical task data (estimated hours, actual hours, AI utilization, task type) b) Calculates team-specific acceleration factors per task type c) Provides confidence intervals for future estimates d) Generates a report showing estimation accuracy trends over time e) Flags tasks where estimates are consistently inaccurate

Exercise 29: Stakeholder Dashboard Design

Design a project dashboard (describe the layout, metrics, and visualizations) specifically tailored for AI-augmented project reporting. Your dashboard should include: a) At least six metrics relevant to AI-augmented development b) Two different views (one for technical stakeholders, one for business stakeholders) c) Visual indicators for AI-specific risks d) A burndown chart adapted for asymmetric acceleration e) A section for AI tool health and utilization

Exercise 30: Full Project Plan

Create a complete project plan for the following scenario: A team of four developers (two experienced with AI tools, two new to them) must build a customer relationship management (CRM) system in 10 weeks. The CRM requires: - Contact management with import/export - Deal pipeline with stages and reporting - Email integration (sending and tracking) - Activity logging and reminders - Dashboard with KPI metrics - Role-based access control

Your plan should include: a) Task decomposition with AI acceleration tier classification b) Three-point AI estimates for each major feature c) A sprint-by-sprint schedule (5 two-week sprints) d) Risk register with AI-specific risks e) Stakeholder communication plan with the Three Timelines f) Metrics and tracking plan g) Team skill development plan for the two AI-novice developers

For coding exercise solutions, see code/exercise-solutions.py.