Module 8: Capstone & Beyond

Scrum Methodology — From Agile Foundations to Running Real Sprints

Module 8 of 8 Capstone Project Requires Module 7 ~5 hours

Module Overview

Learning Objectives

By the end of this module, you will be able to:

  1. Plan and simulate two complete, consecutive Sprints end-to-end without scaffolding, using real Sprint 1 data to drive decisions
  2. Incorporate Sprint Review feedback and Retrospective actions into subsequent Sprints — closing the empirical loop
  3. Forecast velocity before a Sprint, compare forecast to actual, and explain the gap honestly
  4. Write a release-readiness assessment: what stands between "increment done" and "product shipped"
  5. Compare Nexus, SAFe, and LeSS, and explain when (and whether) each is appropriate
  6. Describe how Kanban flow metrics strengthen Scrum without replacing it
  7. Choose a certification path (PSM I, CSM, PSPO I, PMI-ACP) that fits your goals and budget
  8. Answer the ten most common Scrum interview questions with evidence from your own TaskFlow sprints
Course Information
  • Module: 8 of 8 — the final module
  • Format: One large graded capstone (most of your ~5 hours) plus career-oriented theory
  • Prerequisites: Module 7 completed — you need your TaskFlow Jira project (key TF), your Sprint 1 velocity of ~12–13 points, your Definition of Done, the two review feedback items ("task priorities/colors" and "delete confirmation"), and your retrospective action
  • Tools needed: Your existing Jira TaskFlow project and a document editor for the evidence pack

How this module works

This module is different from the previous seven. There is one big graded capstone — delivering TaskFlow Sprints 2 and 3 in Jira as the full Scrum team — and then a look beyond single-team Scrum: scaling frameworks, Scrum-with-Kanban, certifications, and interview preparation. The capstone proves that Modules 1–7 stuck; the "beyond" sections send you into the real world with honest advice rather than marketing.

Capstone: Deliver TaskFlow Sprints 2 & 3

In Module 7 you ran Sprint 1 with the course guiding every step. Now the training wheels come off. You will plan and simulate two more complete Sprints of TaskFlow, and every decision — the goals, the forecasts, the mid-sprint responses, the retro format — is yours to make and defend. Treat this like a brief from a real employer, because that is exactly what it rehearses.

The Brief

Capstone Brief

Role: You act as the entire Scrum team — Product Owner ordering the backlog, Developers forecasting and executing, Scrum Master keeping the process honest. Where the accountabilities would disagree in real life, write down the tension and how you resolved it.

Mission: Plan and simulate two complete two-week Sprints (Sprint 2 and Sprint 3) in your existing Jira project, picking up exactly where Sprint 1 ended.

Continuity requirements — Sprint 1's outcomes are not decoration:

  • The two Sprint 1 review feedback items — "task priorities/colors" and "delete confirmation" — must become proper user stories with acceptance criteria, refined, estimated, and ordered into the Product Backlog. Whether they land in Sprint 2 or 3 is a Product Owner decision you must justify.
  • Your Sprint 1 retrospective action must be visibly applied. If it was "write acceptance criteria before planning", your Sprint 2 planning notes must show it happening. An improvement that exists only in the retro notes does not count.
  • Your Definition of Done stays in force. If you amend it, record the amendment and the reason.

Ground rules: Simulated does not mean sloppy — issues transition through the board day by day, the daily log is written as-you-go (not reconstructed at the end), and every event produces an artifact.

Sprint 2 Requirements

Sprint 2 — Requirements
1. Sprint Planning
  • Write a Sprint Goal — one sentence describing an outcome, not a list of tickets.
  • Set capacity from evidence: Sprint 1 velocity was ~12–13 points. Forecast accordingly and state your reasoning.
  • Include at least one story from the Collaboration epic — stakeholders are asking when teams can share projects.
2. One scripted mid-sprint event
  • Invent either a blocker or a scope challenge, place it on a specific day, and resolve it correctly. Examples: a third-party API dependency stops a story on day 3 (blocker); a stakeholder demands a new report feature on day 6 (scope challenge — recall the Day-4 dark-mode scenario from Module 7).
  • The resolution must be procedurally correct: blockers surface at the Daily Scrum and are escalated by the Scrum Master; scope challenges go through the Product Owner, who negotiates with the Developers if the Sprint Goal is at risk — and the goal is never quietly rewritten.
3. Full daily log — 10 entries
  • One entry per working day: what moved on the board, the plan for the next 24 hours, any impediments. Your scripted event must appear on the day it happens.
4. Sprint Review
  • Demonstrate the increment against the Sprint Goal and invent two new stakeholder feedback items — plausible ones that follow from what was shown. They join the Product Backlog like the Sprint 1 items did.
5. Sprint Retrospective — new format
  • Use a different format than Start-Stop-Continue. Pick one: 4Ls (Liked, Learned, Lacked, Longed for), Mad-Sad-Glad, or the sailboat (wind, anchors, rocks, island). Produce at least one concrete, testable improvement action for Sprint 3.

Sprint 3 Requirements

Sprint 3 — Requirements

Sprint 3 demands the same rigor as Sprint 2 — goal, evidence-based capacity, ten daily log entries, a review with feedback, a retro with an action — plus two additions:

1. Velocity forecast — BEFORE you start
  • Before planning Sprint 3, write down your predicted velocity and the reasoning (you now have two data points). Seal it — no editing after the fact.
  • After Sprint 3 closes, compare prediction to actual and explain the difference honestly. "We predicted 14 and delivered 11 because the blocker cost us a 3-point story" is a good answer; "we were basically right" with no numbers is not.
2. Release-readiness assessment
  • At the end of Sprint 3, answer: what would it take to ship TaskFlow v1.0 to real users? Which epics are done, partial, or untouched? Is the core workflow (create, edit, complete, delete tasks) truly usable end-to-end? What non-functional work — security, performance, accessibility, hosting — sits outside your Definition of Done? Is the DoD itself strong enough for production?

Deliverables Checklist

Evidence Pack — Submit All of the Following
  • Jira sprint reports ×2 — completed-sprint reports for Sprints 2 and 3 (committed vs. completed, scope changes visible)
  • Burndown readings ×2 — each sprint's burndown with a 2–3 sentence interpretation of its shape
  • Review notes ×2 — what was demonstrated, feedback captured, resulting backlog items
  • Retro outputs ×2 — the full board from each format used, plus the action and evidence it was applied
  • Velocity comparison table — Sprints 1–3: forecast, committed, completed, and your Sprint 3 prediction vs. actual
  • Release recommendation — half a page: ship v1.0 now, after N more sprints, or not yet — with reasons a non-technical stakeholder could follow
The most common capstone failure

Writing the whole daily log in one sitting at the end. It always shows: entries are too tidy, nothing goes wrong until the scripted event, and the Jira board history contradicts the log. Simulate day by day — even at ten minutes per "day", the artifacts come out authentic.

Capstone Grading Rubric

The capstone is graded out of 100. Notice where the weight sits: a quarter of the grade is in-sprint discipline — the day-by-day work that separates people who can talk about Scrum from people who can run it.

Component What we look for Points
Backlog stewardshipSprint 1 feedback items refined into real stories with acceptance criteria, estimated, deliberately ordered; new review items handled the same way15
Sprint planning qualityOutcome-focused Sprint Goals; capacity derived from velocity evidence, not wishes; Collaboration epic story in Sprint 215
In-sprint discipline & event artifacts20 authentic daily log entries; board history consistent with the log; scripted event resolved procedurally correctly; retro action visibly applied25
Review & retro qualityReviews inspect the increment against the goal and produce plausible feedback; retro uses a new format and yields a testable action15
Velocity analysis & forecastSealed Sprint 3 prediction with reasoning; honest post-sprint comparison; complete three-sprint table10
Release recommendationGrounded in the actual state of the four epics and the DoD; a clear, defensible ship/don't-ship call10
Professionalism of evidence packComplete, organised, labelled, readable by someone who has never seen your project10
Total100

What "excellent" looks like

An excellent capstone reads like the record of a team that actually existed: velocity wobbles believably, the blocker ripples into the burndown, the Sprint 2 retro action changes something observable in Sprint 3, and the release recommendation admits what is not done. Perfection is a red flag; honesty is the rubric's spine.

Scaling Scrum: When One Team Isn't Enough

Everything in this course assumed one team, one product, one backlog — and that is where Scrum shines. But products outgrow one team: TaskFlow with a million users might need web, mobile, and platform teams all touching the same product. The moment multiple teams share a product, problems appear that the Scrum Guide does not address: integration (do the increments combine into one working product?), dependencies (team A blocked on team B), and coordination (whose backlog wins?). Scaling frameworks exist to answer these. Here are the three you will hear about, described honestly.

Nexus (Scrum.org)

Nexus is Scrum.org's own answer, designed to be the minimum addition that makes 3–9 teams work on one product. There is still a single Product Owner and a single Product Backlog — non-negotiable, because two backlogs means two products. A Nexus Integration Team, drawn from the Scrum teams, is accountable for the combined work integrating into one increment every Sprint. Events get scaled counterparts (Nexus Sprint Planning, a Nexus Daily Scrum focused on integration, Nexus Sprint Review, Nexus Retrospective). If you know Scrum, you can read the entire Nexus Guide in half an hour — it is deliberately thin.

SAFe (Scaled Agile Framework)

SAFe targets the enterprise: hundreds or thousands of people, multiple products, budgeting cycles, compliance. Its central construct is the Agile Release Train (ART) — a "team of teams" of roughly 50–125 people that plans together in a two-day PI Planning event covering the next 8–12 weeks. SAFe adds many roles (Release Train Engineer, Product Management, System Architect, Business Owners) and several configuration layers.

The controversy, stated plainly: SAFe is the most widely adopted scaling framework in large corporations — and the most criticised. Many experienced agilists consider it too heavy, arguing it re-installs the top-down planning, fixed roles, and big-batch thinking agile was created to escape, wrapped in agile vocabulary. Defenders reply that enterprises need some structure to change at all. Both can be true. For your career: SAFe jobs are plentiful, and you should walk into one with single-team fundamentals intact rather than learning agile through SAFe.

LeSS (Large-Scale Scrum)

LeSS — the name is a pun, "LeSS is more" — takes the opposite bet: scaling means removing, not adding. Created by Craig Larman and Bas Vodde, LeSS is essentially the Scrum Guide applied to 2–8 teams with almost nothing added: one Product Owner, one Product Backlog, one Sprint for all teams, one integrated increment. Coordination happens between teams directly ("just talk"), not through new roles or layers. LeSS is intellectually elegant and organisationally demanding — it usually requires restructuring the organisation itself, which is why fewer companies adopt it than SAFe.

NexusSAFeLeSS
Teams supported3–9 teams, one product50–125 people per ART; whole enterprises via portfolios2–8 teams (LeSS Huge: more, via requirement areas)
Added rolesNexus Integration TeamMany: RTE, Product Management, System Architect, Business OwnersNone (LeSS Huge adds Area Product Owners)
Added eventsNexus-level planning, daily, review, retroPI Planning, ART syncs, System Demo, Inspect & AdaptSlight tweaks: shared planning part 1, overall retro
WeightLight — a thin layer on ScrumHeavy — layers, roles, its own certification ecosystemMinimal — deliberately less process, not more
Master single-team Scrum first

Every scaling framework assumes the teams inside it already run excellent single-team Scrum. Scaling a broken process gives you a bigger broken process — dysfunction multiplied by the number of teams. If an employer drops you into a Nexus or an ART, the skills that make you valuable there are the ones you practised here: a clean backlog, honest forecasting, a real Definition of Done, and events that actually inspect and adapt. Learn to scale after the fundamentals are boring to you.

Scrum with Kanban

Module 1 presented Scrum and Kanban as alternatives. In practice, mature teams often run Kanban's flow practices inside Scrum — the Sprint, roles, and events stay, and the team adds flow optimisation on top. This is common enough that Scrum.org publishes an official Kanban Guide for Scrum Teams (see Resources) describing exactly this combination. The three additions that matter most:

WIP limits

Cap how many items may sit in "In Progress" on the sprint board (e.g., max 3). Forces the team to finish before starting — the cure for ten stories all 80% done on day 9. "Stop starting, start finishing."

Cycle time

How long an item takes from work starting to done. If average cycle time is 2 days, a story picked up on day 9 of a 10-day sprint is a visible risk — no burndown squinting required. Jira tracks this in its control chart.

Throughput

How many items the team finishes per unit of time (e.g., 5 items/week). Velocity weighs items by size; throughput just counts them. Together they make forecasting more robust — especially when estimates are noisy.

When should a team consider it?

The clearest signal is lots of unplanned work. Imagine TaskFlow is live and your team spends 40% of each sprint on urgent bugs and support escalations that cannot wait two weeks. Pure Scrum handles this awkwardly — unplanned work either torpedoes the forecast or queues painfully in the backlog. Flow metrics help: WIP limits keep interruptions from swamping planned work, cycle time tells you how fast you really turn around an urgent item, and throughput gives stakeholders an honest service-level expectation ("urgent bugs are typically resolved within 3 days"). Other signals: work that resists estimation, or a burndown that routinely cliff-drops on the last day because everything finishes at once.

It is still Scrum

The Kanban Guide for Scrum Teams is explicit: nothing in Scrum is removed. The Sprint Goal, events, accountabilities, and Definition of Done all remain. Kanban practices are an overlay that improves transparency of flow — one more example of the Module 2 lesson that Scrum is a minimal framework designed to have practices added inside it.

Certification Paths

Scrum certifications are a genuine hiring signal — many job postings filter on them — but the market is confusing because two rival bodies (Scrum.org and Scrum Alliance) offer similar-sounding credentials. The honest comparison:

PSM I — Professional Scrum Master I
  • Body: Scrum.org (founded by Ken Schwaber, Scrum's co-creator)
  • Cost: ~$200 per exam attempt
  • Format: Online, anytime — 80 questions, 60 minutes, 85% to pass
  • Course required: No — self-study fully accepted
  • Validity: Lifetime, no renewal fees ever
  • Reputation: Rigorous exam, tightly aligned to the Scrum Guide
CSM — Certified ScrumMaster
  • Body: Scrum Alliance (the older organisation)
  • Cost: mandatory 2-day course, typically ~$500–1000 (exam included)
  • Format: Exam after the course — widely considered easy after attending
  • Course required: Yes — no self-study route exists
  • Validity: Renewal every 2 years (fee + education credits)
  • Reputation: Long-established brand; the live course is the real product
PSPO I — Professional Scrum Product Owner I
  • Body: Scrum.org — same format, cost, and lifetime validity as PSM I
  • Focus: The Product Owner path — value maximisation, backlog management, stakeholder work
  • For whom: If Modules 3–4 (the PO accountability, backlog craft, ordering) were your favourite part of this course, this is your track
  • Note: Many practitioners take both PSM I and PSPO I — the overlap is large
PMI-ACP — Agile Certified Practitioner
  • Body: Project Management Institute (of PMP fame)
  • Scope: Broader than Scrum — Kanban, XP, Lean, general agile practice
  • Barrier: Requires documented agile work experience hours plus training — not realistic straight out of a course
  • For whom: Professionals a few years in, especially in PMI-oriented organisations

Honest advice for students

PSM I is the strongest value for someone finishing this course: a fraction of a CSM's cost, no travel or scheduling, never expires, and its difficulty makes it a more credible signal, not less. The best preparation is free: the Scrum Guide itself plus the scrum.org open assessments. You do not need a paid prep course; you need repetitions.

Suggested PSM I Study Plan
  1. Read the Scrum Guide three times. First: cover to cover (it is 13 pages). Second: annotate — every sentence answers an exam question. Third: read it adversarially, asking "what does this not say?" — many exam traps are common practices (velocity, story points, burndowns) the Guide never actually mandates.
  2. Take the Scrum Open assessment on scrum.org repeatedly until you score a consistent 100% — three clean runs in a row, not one lucky pass.
  3. Review Modules 2–5 of this course — the framework, accountabilities, events, and artifacts map directly onto the exam blueprint.
  4. Sit the exam with the clock in mind: ~45 seconds per question; flag and return rather than stall.

Scrum in Job Interviews

Whether you interview as a developer, Scrum Master, or product person, Scrum questions will come up — and interviewers listen for understanding, not recitation. The ten questions below cover most of what gets asked; each sketch shows the shape of a strong answer. Make it yours by substituting evidence from your own TaskFlow sprints.

Strong answer: Not "runs the meetings" — that is the weak answer. A Scrum Master is accountable for the team's effectiveness: removing impediments, coaching self-management, protecting focus, helping the Product Owner with backlog techniques, and helping the wider organisation understand empirical working. On a good day the events almost run themselves and the Scrum Master works one level up — on the system around the team.

Strong answer: Route it through the Product Owner — stakeholders never inject work directly into a Sprint. Then it is a conversation, not a flat "no": if the request endangers the Sprint Goal, it waits for next planning; if genuinely urgent, the PO and Developers can swap out similar-sized scope as long as the goal survives. You have lived this: in Module 7's Day-4 dark-mode scenario, the correct play was PO triage — capture, size, order into the backlog, defend the running Sprint. Tell that story; concrete beats abstract.

Strong answer: Points estimate relative size and complexity, not duration. Humans are poor at absolute estimates but decent at comparison ("this is about twice that"). Points stay valid across people — a 5 is a 5 whoever picks it up, while "6 hours" depends on who is asking — and they feed velocity, which converts relative estimates into an empirical forecast. The caveat that earns respect: points are a team-internal planning tool, not a productivity metric — and the Scrum Guide does not actually mandate them at all.

Strong answer: The Sprint does not end early — its length is fixed — but the team is not idle: the Developers pull the next-ordered backlog item in collaboration with the Product Owner, or invest in refinement, tests, or technical debt. If it becomes a pattern, that is a planning signal for the Retrospective: forecast more. The trap answers are "extend the scope silently" or "end the sprint" — both break the cadence that makes velocity meaningful.

Strong answer: They inspect different things. The Review inspects the product: team and stakeholders examine the increment against the Sprint Goal and adapt the Product Backlog — a working session, not demo theatre. The Retrospective inspects the process: the Scrum Team alone examines how the Sprint went and commits to an improvement. One adjusts what we build; the other, how. In your TaskFlow Sprint 1, the review produced two backlog items and the retro produced a process action — perfect one-line evidence.

Strong answer: Yes — nothing in the Scrum Guide requires co-location, and a large share of Scrum teams are now remote or hybrid. But remote raises the cost of the informal communication Scrum leans on, so good remote teams compensate deliberately: a genuinely current digital board as the single source of truth, tight facilitation in events, more written artifacts (working agreements, decision logs), and protected overlap hours for the Daily Scrum. The honest caveat: transparency takes more discipline remotely — the board is the room now.

Strong answer: Velocity is the work (usually story points) a team completes per Sprint, used by that team to forecast capacity. And no — you cannot compare it across teams. This is a favourite trap. Points are relative units calibrated inside each team: Team A's 8 might be Team B's 3, because they anchored on different reference stories. Comparing velocities is comparing measurements taken with different rulers — and once velocity becomes a management scoreboard, teams inflate estimates and the metric destroys itself. What can be compared: outcomes and value delivered, not output points.

Strong answer: Interviewers love this one because zealots fail it. Scrum fits complex product work where requirements emerge through feedback. It fits poorly when the work is interrupt-driven (support/ops — Kanban's continuous flow fits better); truly repeatable and fully specified (a fixed regulatory port with zero discovery); when there is no empowered Product Owner or no stakeholder access (empiricism starves without feedback); or when the "team" is one person (the events collapse into overhead). Knowing a tool's boundaries is evidence you understand the tool.

Strong answer: This is where this course pays off — you have real sprints to narrate. Structure: goal → plan → what happened → what we learned. "Our Sprint Goal was a usable end-to-end task workflow. We forecast 13 points from a velocity of 12. On day 4 a stakeholder pushed dark mode mid-sprint; the PO captured it into the backlog and the goal survived. We completed 12 points, the review produced two feedback items — task priorities and a delete confirmation — and our retro action was verified the next sprint." Specific numbers, one complication, one improvement: that sounds senior at any seniority.

Strong answer: A checklist of quality criteria every increment must meet — objective, checkable, applied to all work, never renegotiated per story (that is what acceptance criteria are for). A strong one covers: code reviewed, tests passing, integrated with everyone else's work, documented where needed, deployable. Two marks of maturity: the DoD should grow stricter as the team's capability grows, and "done" must mean genuinely usable — a DoD permitting "done but not tested" manufactures hidden work and destroys the transparency empiricism depends on.

The meta-skill

Notice the pattern across all ten sketches: rule → reason → real example. State what Scrum says, explain the mechanism behind it, then ground it in a sprint you actually ran. Candidates who can only do the first step sound like flashcards; you can now do all three.

Course Wrap-up

Eight modules ago, TaskFlow was a wish list and "agile" was a buzzword. Look at what you can do now:

What You Can Now Do
  • Module 1: Explain why agile exists — waterfall's failure modes, the Manifesto's values and principles, and when Scrum, Kanban, or XP fits
  • Module 2: Describe the complete Scrum framework — empiricism, the Scrum values, and how every element connects
  • Module 3: Distinguish the three accountabilities and spot the anti-patterns
  • Module 4: Build and manage a real Product Backlog in Jira — epics, user stories, acceptance criteria, ordering by value
  • Module 5: Run all five Scrum events with purpose — planning that produces a goal, dailies that replan, reviews that inspect, retros that change something
  • Module 6: Estimate with story points, read burndowns, forecast with velocity — and know the limits of every metric
  • Module 7: Execute a full Sprint end-to-end in Jira, handling mid-sprint pressure without breaking the framework
  • Module 8: Run consecutive sprints independently, forecast and verify, assess release readiness — and see over the horizon to scaled and hybrid Scrum

Where to Go Next

XP Practices

Scrum says nothing about how to build quality software inside the Sprint. Extreme Programming does: test-driven development, pair programming, refactoring, continuous integration. The strongest Scrum teams are quietly XP teams inside their Sprints — start with TDD and your Definition of Done grows teeth.

DevOps & CI/CD

A "done" increment every Sprint begs the question: why not ship every Sprint — or every day? Continuous integration and delivery pipelines close the gap between increment and production, and pair naturally with everything you learned here.

DevOps Lab course
Product Management

If ordering the TaskFlow backlog was the part you loved — deciding what to build and why — the product path awaits: discovery, user research, outcome metrics, roadmapping. The PSPO track is the certification on-ramp; Inspired (Marty Cagan) is the reading on-ramp.

Resources & References

Practice & Community
Course Complete
Congratulations — you made it.

You started this course with a wish list on a whiteboard and you are finishing it with three simulated Sprints of a real product in Jira, a velocity history, a Definition of Done you wrote yourself, and a release recommendation with your name on it. That is not "learned about Scrum" — that is ran Scrum. Whatever you build next, build it in small slices, show it to real people early, and change course when they teach you something. That habit is the whole course in one sentence.

Back to Course Home