A $24 million product group has 10 teams and 96 people working on three products over a shared platform, uses two-week iterations, and funds work quarterly for the next 18 months. Which scaling approach would you choose without importing an entire framework by default?

I would identify the development value streams before choosing a scaling model because a shared platform alone does not prove that all 10 teams belong in one ART. - Product and Engineering leaders would map how each product moves from demand to released value, including customers, release cadence, system boundaries, handoffs, and recurring dependencies. - If all 10 teams must collaborate to release one integrated solution, I would test one ART; if the three products release independently, I would keep separate product groups and treat the platform through explicit service and dependency policies. - I would record the decision after a two-week mapping workshop and review it after two planning cycles against dependency age, integrated-delivery frequency, and coordination cost capped at 8% of capacity.

Six feature teams with 54 people build one commerce product from one Product Backlog, run two-week Sprints, and must cut feature lead time from 42 to 25 days within nine months on a $7 million roadmap. Would you use LeSS, SAFe, Nexus, or lightweight coordination?

I would use LeSS because one product, one Product Owner, one Product Backlog, and six feature teams fit its operating model. - The Product Owner and I would maintain a LeSS adoption canvas that names the product boundary, team redesign, one Sprint cadence, Overall Product Backlog Refinement, Sprint Review, and Overall Retrospective. - I would budget four weeks for multiteam feature-team coaching and remove separate program roles, while the Product Owner remains accountable for ordering the single Product Backlog. - We continue after three months only if at least 70% of backlog items finish within one team and the 85th-percentile feature lead time is on a path to 25 days.

Five Scrum Teams with 45 people deliver one identity platform in two-week Sprints, but 40 cross-team integration defects appear each quarter and the next product release is six months away. Which scaling framework would you introduce and what would it cost?

I would introduce Nexus because the immediate constraint is producing one integrated Increment across five Scrum Teams every Sprint. - I would create a Nexus operating agreement owned by the Nexus Integration Team that covers cross-team refinement, the Nexus Sprint Backlog, Nexus Daily Scrum, integrated testing, and one Nexus Sprint Review. - The Nexus Integration Team, including the Product Owner, a Scrum Master, and appropriate members, is accountable for enabling a valuable integrated Increment; it does not become a separate delivery team or assign work. - I would cap setup at 180 person-hours and exit the intervention after six Sprints when integration defects stay below 10 per quarter and one integrated Increment is Done every Sprint.

Four autonomous product teams with 36 people deploy weekly, share only two dependencies per month, plan in two-week Sprints, and have a six-month $3 million growth horizon. Would you add a scaling framework?

I would not add one because the dependency load does not justify permanent scaled roles or events. - I would own a one-page coordination policy with the four Product Owners that defines a monthly product-group review, an asynchronous dependency log, and escalation only for items older than five working days. - Each team keeps its own Scrum accountabilities and Product Backlog, while a rotating Scrum Master facilitates the 45-minute monthly review at a cost below 12 person-hours per month. - We reconsider Nexus, LeSS, or an ART only if shared dependencies exceed eight per month, integrated releases become necessary, or dependency age exceeds 10 days for two months.

Eight teams with 78 people are building one regulated medical product on two-week iterations, face quarterly evidence reviews, and have 12 months and $18 million to reach approval. How would you adapt SAFe without turning compliance into extra ceremony?

I would use an Essential SAFe ART with compliance work built into the Definition of Done and the ART Backlog rather than create a parallel waterfall track. - The Release Train Engineer would own the 10-week ART calendar, while Quality and Regulatory own an evidence map linking each compliance control to features, enablers, tests, and review dates. - Product Management orders the ART Backlog and System Architect guides the architectural runway; Scrum Masters coach teams to produce evidence within each two-week iteration. - We retain each event only when it produces traceable evidence or a decision, with approval readiness above 95% and coordination cost below 10% of the $18 million delivery budget.

An eight-team ART depends on four security and data specialists who can each give 30% to 50% of their time, and it also consumes a deployment platform run by a dedicated product team. How would you represent these two kinds of support in PI Planning?

I would represent the scarce specialists as Shared Services on the ART while keeping the deployment platform as a separate product with its own backlog and service boundary. - Shared Services representatives join the relevant PI Planning breakouts, publish capacity by iteration, and place dated specialist dependencies on the ART planning board instead of appearing as a fictional full-time team. - Product Management orders ART Backlog Features and Product Owners order Team Backlogs, while the platform product manager orders platform demand against published service levels and technical strategy. - The RTE facilitates capacity and dependency visibility, but platform requests stay on the platform backlog; the model is reviewed when specialist over-allocation exceeds 10% or platform request lead time exceeds 10 working days.

Ten feature teams with 92 people work on one global marketplace, deploy weekly, use two-week Sprints, and must halve a 46-day idea-to-market time within 12 months on a $16 million budget. Would you choose LeSS Huge or Essential SAFe?

I would choose LeSS Huge if leadership accepts one product, one overall Product Owner, and feature teams organized around customer value. - The overall Product Owner and I would define two customer-centered Requirement Areas with five teams in each, keeping every area within the LeSS Huge range of four to eight teams and retaining one Product Backlog. - Area Product Owners would specialize in refinement within their areas without replacing the overall Product Owner, while Scrum Masters coach feature-team adoption and the Overall Retrospective. - I would cap the first two quarters of redesign and coaching at $600,000 and continue only if cross-area handoffs fall 40% and 85th-percentile idea-to-market time falls below 30 days.

Nine teams with 84 people combine hardware, firmware, and cloud work, integrate every 12 weeks, use two-week team iterations, and have a $25 million 18-month product horizon. Which coordination model would you select?

I would use one Essential SAFe ART with a 12-week Planning Interval because synchronized integration and fixed hardware lead times dominate the coordination need. - The Release Train Engineer would maintain the ART event calendar and planning board, while Product Management owns Features and the System Architect owns interface and runway decisions. - The ART would hold a System Demo after every two-week iteration, while a separate four-week milestone review tracks supplier and hardware dates; teams would still plan their own iteration work. - The model passes a two-PI review only if at least 80% of dated milestones are hit within their agreed windows, interface dependency age stays below 14 days, and ART overhead remains below $1.5 million annually.

Nine Scrum Teams deliver three independently releasable SaaS products under one strategy, but shared identity and data dependencies create 18 unresolved cross-product decisions each month. Leadership proposes Nexus even though there is no single Product Backlog or integrated Increment. What would you use?

I would pilot a minimal Scrum@Scale coordination pattern rather than Nexus because these teams need a network across products, not one Nexus producing a single integrated Increment. - A Scrum of Scrums would meet twice weekly around dependency age and impediments, with a Scrum of Scrums Master facilitating escalation rather than assigning work to the nine teams. - An Executive MetaScrum would bring the three Product Owners and decision-making stakeholders together to sequence shared identity and data demand while each Product Owner retains the backlog for an independently releasable product. - I would keep the pilot for eight weeks only if unresolved decisions fall from 18 to fewer than six per month and coordination stays below 60 person-hours monthly; the cost is two added decision forums and explicit scaled roles.

Three ARTs with 24 teams and an external signaling supplier must deliver one integrated rail-control solution every six months, and no ART can validate the customer outcome alone. When is a Solution Train actually justified?

I would justify a Solution Train only if the integrated solution requires persistent coordination across multiple ARTs and suppliers that cannot be contained within one ART. - The evidence must show one solution boundary, shared Solution Intent, cross-ART Capabilities, interface milestones, and integration decisions whose economic impact spans all three ARTs. - If the work can be reorganized into one ART or decoupled through stable interfaces, I would use those cheaper options; a large budget or many teams by itself is not sufficient. - Once the criteria hold, Solution Management, a Solution Architect, and a Solution Train Engineer take their distinct accountabilities, with the extra layer reviewed after two increments against integration delay and coordination cost.

Eight teams with 76 people across two time zones are entering a 10-week Planning Interval, work in two-week iterations, and must deliver a $6 million customer outcome in six months. How would you design PI Planning?

I would run a two-day PI Planning event with one shared outcome, protected team breakouts, and immediate resolution of cross-team constraints. - The Release Train Engineer owns the agenda and readiness checklist, Product Management brings vision and prioritized features, and Business Owners bring business context and decision authority. - Each team produces a capacity-based plan and draft PI Objectives, then the ART creates a planning board, ROAMs program risks, and revises plans after the management review and problem-solving session. - Planning exits only with 100% of teams represented, every critical dependency owned, no unowned risk, and a final confidence vote of at least 3 out of 5 from every team.

Six teams with 54 people plan an eight-week PI in two-week iterations, have 30 candidate features, and need four measurable product outcomes within a $5 million half-year roadmap. How would you produce useful PI Objectives?

I would make teams express PI Objectives as measurable business or technical outcomes rather than restated feature lists. - Product Management owns the ordered ART Backlog, while each team owns its PI Objectives and records committed and uncommitted objectives on a shared objective sheet. - Business Owners assign planned business value after discussing outcomes, and the Release Train Engineer facilitates consistency without writing or scoring objectives for them. - The artifact is complete when all 30 features map to no more than four outcome themes, every objective has a measure, and no team carries work above 85% planned capacity.

Nine teams with 86 people use two-week iterations inside a 10-week PI, reserve between 5% and 30% for support, and have a $9 million release due in five months. How would you make capacity planning comparable without standardizing velocity?

I would standardize capacity assumptions and risk categories, not story points or team velocity. - Each team owns a capacity sheet using its available people-days, planned leave, support allocation, and historical unplanned-work range for the five iterations. - Product Management uses those sheets to sequence features, while the Release Train Engineer exposes aggregate capacity and the Scrum Masters challenge hidden overload in team breakouts. - Planning exits when every team stays at or below 85% allocatable capacity, support reserves reflect the last three PIs, and no feature depends on cross-team velocity conversion.

Seven teams with 64 people have 48 planned dependencies across two-week iterations and a five-month release horizon, with current dependency age at 18 days. How would you build and operate the PI dependency board?

I would make the ART planning board a forecast of feature dates and cross-team dependencies, not a decorative wall after PI Planning. - During planning, the delivering and receiving teams co-own each dependency card with needed-by iteration, acceptance condition, named contacts, and a link to the relevant feature. - The Release Train Engineer owns board hygiene and reviews aging weekly in ART Sync, while teams update status and Product Management resolves sequencing choices. - The board is usable when all 48 dependencies have owners, zero critical links miss a needed-by date, and 85th-percentile dependency age falls from 18 to under 10 days by PI end.

Eight teams with 80 people finish two-day PI Planning for a 10-week interval, but the first confidence vote averages 2.8 out of 5 and the roadmap commits $7 million over six months. What should happen before planning closes?

I would keep planning open, surface the low votes, and revise the plan before asking for a second confidence vote. - The Release Train Engineer owns an anonymous concern log, invites every voter below 3 to name the constraint, and routes scope issues to Product Management and capacity issues back to teams. - Teams update PI Objectives, dependencies, and risks after the management review and problem-solving session; Business Owners clarify priorities without pressuring scores upward. - PI Planning closes only when every team vote averages at least 3, all material concerns have an owner, and removed scope is visible in the final planning board and objective set.

Ten teams with 100 people work in two-week iterations, carry 35 active dependencies, meet on a 10-week PI cadence, and have a nine-month delivery horizon. How would you design ART Sync?

I would run a weekly 45-minute ART Sync that combines the Coach Sync and PO Sync views around flow, scope, and impediments. - The Release Train Engineer owns a decision-focused agenda fed by the planning board, dependency aging report, PI Objective status, and top ART impediments. - Scrum Masters and Team Coaches bring flow and impediment data, while Product Owners and Product Management bring scope and priority decisions; the RTE facilitates but does not decide backlog order. - The event remains weekly while decisions close within three working days, critical dependency age stays below 10 days, and follow-up work consumes under 100 person-hours per PI.

Seven teams with 68 people build an integrated service in two-week iterations, plan every 10 weeks, and have a $10 million release in eight months. How would you design the System Demo?

I would demonstrate the integrated Increment from a production-like environment at the end of every iteration, not present separate team slide decks. - The System Team owns the demo environment and coordinates integration readiness, teams remain responsible for working slices, and the Release Train Engineer owns the 60-minute facilitation plan. - Product Management frames progress against PI Objectives and captures stakeholder feedback in the ART Backlog, while System Architect records architectural findings without turning the demo into a gate. - The demo succeeds when all seven teams contribute to one integrated flow, at least 90% of planned slices run live, and accepted feedback receives an owner within two working days.

Nine teams with 90 people close a 10-week PI after five two-week iterations, have a 24-day median feature lead time, and need to reach 16 days within 12 months. How would you run Inspect and Adapt?

I would run the full Inspect and Adapt event: PI System Demo, quantitative and qualitative review, then a focused problem-solving workshop. - The Release Train Engineer owns the event record, Business Owners provide actual business value, and Product Management connects delivered outcomes to the next ART Backlog choices. - Teams use flow data and a root-cause artifact to select one systemic problem, then create improvement backlog items with named owners rather than a broad action list. - The event exits with no more than three funded improvements, each due within the next PI, and feature lead time must trend from 24 toward 16 days without comparing team velocities.

Six teams with 58 people span five time zones, use two-week iterations and a 10-week PI, and have a $400,000 annual coordination budget. How would you make distributed PI Planning inclusive and affordable?

I would keep one two-day PI Planning outcome while splitting live work into two four-hour overlap windows and facilitated local breakouts. - The Release Train Engineer owns a timezone map, readiness pack, recorded context, Miro planning board, and decision log published 48 hours before the event. - Local Scrum Masters facilitate team breakouts, Product Management and Business Owners attend both decision windows, and asynchronous questions carry a four-hour response service level. - The design passes when all six teams complete objectives and dependencies, participation exceeds 90%, no region works outside 07:00 to 20:00 local time, and annual cost stays below $400,000.

Eight teams with 75 people use two-week iterations, a 10-week PI, and a six-month $8 million roadmap, but leadership wants one coherent ART event calendar before funding. What cadence would you propose?

I would propose a minimal nested cadence: weekly ART Sync, System Demo each iteration, PI Planning at interval start, and Inspect and Adapt at interval end. - The Release Train Engineer owns a six-month event map showing purpose, required roles, input artifact, output decision, and person-hour budget for every ART event. - Product Management owns backlog and content decisions, Business Owners value PI Objectives, System Team supports integration, and Scrum Masters coach team participation. - Funding proceeds when event overhead is below 8% of ART capacity, every event has a measurable output, and duplicate status meetings are removed within the first PI.

Scrum Master interview questions

100 real questions with model answers and explanations for Senior Scrum Master candidates.

See a Scrum Master resume example →

Practice with flashcards

Spaced repetition · Hunter Pass

Questions

scalingiteration

I would identify the development value streams before choosing a scaling model because a shared platform alone does not prove that all 10 teams belong in one ART.

Product and Engineering leaders would map how each product moves from demand to released value, including customers, release cadence, system boundaries, handoffs, and recurring dependencies.
If all 10 teams must collaborate to release one integrated solution, I would test one ART; if the three products release independently, I would keep separate product groups and treat the platform through explicit service and dependency policies.
I would record the decision after a two-week mapping workshop and review it after two planning cycles against dependency age, integrated-delivery frequency, and coordination cost capped at 8% of capacity.

Why interviewers ask this: The interviewer is testing whether the candidate identifies development value streams before selecting SAFe and does not infer one ART from shared technology.

agileroadmapbacklog

I would use LeSS because one product, one Product Owner, one Product Backlog, and six feature teams fit its operating model.

The Product Owner and I would maintain a LeSS adoption canvas that names the product boundary, team redesign, one Sprint cadence, Overall Product Backlog Refinement, Sprint Review, and Overall Retrospective.
I would budget four weeks for multiteam feature-team coaching and remove separate program roles, while the Product Owner remains accountable for ordering the single Product Backlog.
We continue after three months only if at least 70% of backlog items finish within one team and the 85th-percentile feature lead time is on a path to 25 days.

Why interviewers ask this: A strong answer recognizes the organizational assumptions behind LeSS and avoids adding a program hierarchy to a single-product setting.

agilereleasesscaling

I would introduce Nexus because the immediate constraint is producing one integrated Increment across five Scrum Teams every Sprint.

I would create a Nexus operating agreement owned by the Nexus Integration Team that covers cross-team refinement, the Nexus Sprint Backlog, Nexus Daily Scrum, integrated testing, and one Nexus Sprint Review.
The Nexus Integration Team, including the Product Owner, a Scrum Master, and appropriate members, is accountable for enabling a valuable integrated Increment; it does not become a separate delivery team or assign work.
I would cap setup at 180 person-hours and exit the intervention after six Sprints when integration defects stay below 10 per quarter and one integrated Increment is Done every Sprint.

Why interviewers ask this: The interviewer is checking accurate Nexus accountabilities and whether the candidate ties framework cost to an integration outcome.

agiledeploymentdependencies

I would not add one because the dependency load does not justify permanent scaled roles or events.

I would own a one-page coordination policy with the four Product Owners that defines a monthly product-group review, an asynchronous dependency log, and escalation only for items older than five working days.
Each team keeps its own Scrum accountabilities and Product Backlog, while a rotating Scrum Master facilitates the 45-minute monthly review at a cost below 12 person-hours per month.
We reconsider Nexus, LeSS, or an ART only if shared dependencies exceed eight per month, integrated releases become necessary, or dependency age exceeds 10 days for two months.

Why interviewers ask this: The question evaluates whether the candidate can reject framework cargo cult when a cheap coordination mechanism matches the actual coupling.

iteration

I would use an Essential SAFe ART with compliance work built into the Definition of Done and the ART Backlog rather than create a parallel waterfall track.

The Release Train Engineer would own the 10-week ART calendar, while Quality and Regulatory own an evidence map linking each compliance control to features, enablers, tests, and review dates.
Product Management orders the ART Backlog and System Architect guides the architectural runway; Scrum Masters coach teams to produce evidence within each two-week iteration.
We retain each event only when it produces traceable evidence or a decision, with approval readiness above 95% and coordination cost below 10% of the $18 million delivery budget.

Why interviewers ask this: The interviewer wants a bounded SAFe adaptation that integrates regulatory evidence while keeping role ownership and cost visible.

deployment

I would represent the scarce specialists as Shared Services on the ART while keeping the deployment platform as a separate product with its own backlog and service boundary.

Shared Services representatives join the relevant PI Planning breakouts, publish capacity by iteration, and place dated specialist dependencies on the ART planning board instead of appearing as a fictional full-time team.
Product Management orders ART Backlog Features and Product Owners order Team Backlogs, while the platform product manager orders platform demand against published service levels and technical strategy.
The RTE facilitates capacity and dependency visibility, but platform requests stay on the platform backlog; the model is reviewed when specialist over-allocation exceeds 10% or platform request lead time exceeds 10 working days.

Why interviewers ask this: A strong answer distinguishes part-time Shared Services that support an ART from a platform team that owns an internal product and backlog.

agiledeployment

I would choose LeSS Huge if leadership accepts one product, one overall Product Owner, and feature teams organized around customer value.

The overall Product Owner and I would define two customer-centered Requirement Areas with five teams in each, keeping every area within the LeSS Huge range of four to eight teams and retaining one Product Backlog.
Area Product Owners would specialize in refinement within their areas without replacing the overall Product Owner, while Scrum Masters coach feature-team adoption and the Overall Retrospective.
I would cap the first two quarters of redesign and coaching at $600,000 and continue only if cross-area handoffs fall 40% and 85th-percentile idea-to-market time falls below 30 days.

Why interviewers ask this: The interviewer is testing whether the candidate applies the LeSS Huge Requirement Area structure correctly and makes its organizational preconditions explicit.

iteration

I would use one Essential SAFe ART with a 12-week Planning Interval because synchronized integration and fixed hardware lead times dominate the coordination need.

The Release Train Engineer would maintain the ART event calendar and planning board, while Product Management owns Features and the System Architect owns interface and runway decisions.
The ART would hold a System Demo after every two-week iteration, while a separate four-week milestone review tracks supplier and hardware dates; teams would still plan their own iteration work.
The model passes a two-PI review only if at least 80% of dated milestones are hit within their agreed windows, interface dependency age stays below 14 days, and ART overhead remains below $1.5 million annually.

Why interviewers ask this: The answer must connect SAFe to integration economics, require a System Demo every iteration, and treat the four-week checkpoint as a milestone review measured by milestone hit rate rather than SAFe ART predictability.

agiledependenciescloud

I would pilot a minimal Scrum@Scale coordination pattern rather than Nexus because these teams need a network across products, not one Nexus producing a single integrated Increment.

A Scrum of Scrums would meet twice weekly around dependency age and impediments, with a Scrum of Scrums Master facilitating escalation rather than assigning work to the nine teams.
An Executive MetaScrum would bring the three Product Owners and decision-making stakeholders together to sequence shared identity and data demand while each Product Owner retains the backlog for an independently releasable product.
I would keep the pilot for eight weeks only if unresolved decisions fall from 18 to fewer than six per month and coordination stays below 60 person-hours monthly; the cost is two added decision forums and explicit scaled roles.

Why interviewers ask this: The interviewer is checking whether the candidate can explain a Scrum@Scale trade-off and reject Nexus when its single-product integration premise is absent.

validation

I would justify a Solution Train only if the integrated solution requires persistent coordination across multiple ARTs and suppliers that cannot be contained within one ART.

The evidence must show one solution boundary, shared Solution Intent, cross-ART Capabilities, interface milestones, and integration decisions whose economic impact spans all three ARTs.
If the work can be reorganized into one ART or decoupled through stable interfaces, I would use those cheaper options; a large budget or many teams by itself is not sufficient.
Once the criteria hold, Solution Management, a Solution Architect, and a Solution Train Engineer take their distinct accountabilities, with the extra layer reviewed after two increments against integration delay and coordination cost.

Why interviewers ask this: A senior answer creates a Solution Train for a genuinely large integrated solution, not as an automatic layer above a single ART.

designiteration

I would run a two-day PI Planning event with one shared outcome, protected team breakouts, and immediate resolution of cross-team constraints.

The Release Train Engineer owns the agenda and readiness checklist, Product Management brings vision and prioritized features, and Business Owners bring business context and decision authority.
Each team produces a capacity-based plan and draft PI Objectives, then the ART creates a planning board, ROAMs program risks, and revises plans after the management review and problem-solving session.
Planning exits only with 100% of teams represented, every critical dependency owned, no unowned risk, and a final confidence vote of at least 3 out of 5 from every team.

Why interviewers ask this: The interviewer checks whether the candidate can design PI Planning around its actual inputs, outputs, and role accountabilities.

roadmapiterationroadmapping

I would make teams express PI Objectives as measurable business or technical outcomes rather than restated feature lists.

Product Management owns the ordered ART Backlog, while each team owns its PI Objectives and records committed and uncommitted objectives on a shared objective sheet.
Business Owners assign planned business value after discussing outcomes, and the Release Train Engineer facilitates consistency without writing or scoring objectives for them.
The artifact is complete when all 30 features map to no more than four outcome themes, every objective has a measure, and no team carries work above 85% planned capacity.

Why interviewers ask this: A strong answer distinguishes who writes, prioritizes, and values PI Objectives and keeps them measurable.

releasescapacity-planningcapacity

I would standardize capacity assumptions and risk categories, not story points or team velocity.

Each team owns a capacity sheet using its available people-days, planned leave, support allocation, and historical unplanned-work range for the five iterations.
Product Management uses those sheets to sequence features, while the Release Train Engineer exposes aggregate capacity and the Scrum Masters challenge hidden overload in team breakouts.
Planning exits when every team stays at or below 85% allocatable capacity, support reserves reflect the last three PIs, and no feature depends on cross-team velocity conversion.

Why interviewers ask this: The interviewer is testing safe capacity planning that respects local estimation systems and avoids invalid velocity comparisons.

dependenciesreleasesiteration

I would make the ART planning board a forecast of feature dates and cross-team dependencies, not a decorative wall after PI Planning.

During planning, the delivering and receiving teams co-own each dependency card with needed-by iteration, acceptance condition, named contacts, and a link to the relevant feature.
The Release Train Engineer owns board hygiene and reviews aging weekly in ART Sync, while teams update status and Product Management resolves sequencing choices.
The board is usable when all 48 dependencies have owners, zero critical links miss a needed-by date, and 85th-percentile dependency age falls from 18 to under 10 days by PI end.

Why interviewers ask this: The question evaluates whether the dependency board has bilateral ownership, operating cadence, and a measurable flow purpose.

roadmaproadmapping

I would keep planning open, surface the low votes, and revise the plan before asking for a second confidence vote.

The Release Train Engineer owns an anonymous concern log, invites every voter below 3 to name the constraint, and routes scope issues to Product Management and capacity issues back to teams.
Teams update PI Objectives, dependencies, and risks after the management review and problem-solving session; Business Owners clarify priorities without pressuring scores upward.
PI Planning closes only when every team vote averages at least 3, all material concerns have an owner, and removed scope is visible in the final planning board and objective set.

Why interviewers ask this: The interviewer wants the confidence vote treated as a planning control, not as a morale ritual or approval theater.

designdependenciesiteration

I would run a weekly 45-minute ART Sync that combines the Coach Sync and PO Sync views around flow, scope, and impediments.

The Release Train Engineer owns a decision-focused agenda fed by the planning board, dependency aging report, PI Objective status, and top ART impediments.
Scrum Masters and Team Coaches bring flow and impediment data, while Product Owners and Product Management bring scope and priority decisions; the RTE facilitates but does not decide backlog order.
The event remains weekly while decisions close within three working days, critical dependency age stays below 10 days, and follow-up work consumes under 100 person-hours per PI.

Why interviewers ask this: A strong response distinguishes the two perspectives within ART Sync and protects Product Management and team accountabilities.

system-designdesigniteration

I would demonstrate the integrated Increment from a production-like environment at the end of every iteration, not present separate team slide decks.

The System Team owns the demo environment and coordinates integration readiness, teams remain responsible for working slices, and the Release Train Engineer owns the 60-minute facilitation plan.
Product Management frames progress against PI Objectives and captures stakeholder feedback in the ART Backlog, while System Architect records architectural findings without turning the demo into a gate.
The demo succeeds when all seven teams contribute to one integrated flow, at least 90% of planned slices run live, and accepted feedback receives an owner within two working days.

Why interviewers ask this: The interviewer is testing whether System Demo provides objective integrated evidence and whether supporting roles remain distinct.

iteration

I would run the full Inspect and Adapt event: PI System Demo, quantitative and qualitative review, then a focused problem-solving workshop.

The Release Train Engineer owns the event record, Business Owners provide actual business value, and Product Management connects delivered outcomes to the next ART Backlog choices.
Teams use flow data and a root-cause artifact to select one systemic problem, then create improvement backlog items with named owners rather than a broad action list.
The event exits with no more than three funded improvements, each due within the next PI, and feature lead time must trend from 24 toward 16 days without comparing team velocities.

Why interviewers ask this: The answer should show all Inspect and Adapt components and convert evidence into a small owned improvement backlog.

distributediteration

I would keep one two-day PI Planning outcome while splitting live work into two four-hour overlap windows and facilitated local breakouts.

The Release Train Engineer owns a timezone map, readiness pack, recorded context, Miro planning board, and decision log published 48 hours before the event.
Local Scrum Masters facilitate team breakouts, Product Management and Business Owners attend both decision windows, and asynchronous questions carry a four-hour response service level.
The design passes when all six teams complete objectives and dependencies, participation exceeds 90%, no region works outside 07:00 to 20:00 local time, and annual cost stays below $400,000.

Why interviewers ask this: The interviewer is assessing whether distributed planning preserves shared decisions without imposing hidden timezone and labor costs.

roadmapiterationroadmapping

I would propose a minimal nested cadence: weekly ART Sync, System Demo each iteration, PI Planning at interval start, and Inspect and Adapt at interval end.

The Release Train Engineer owns a six-month event map showing purpose, required roles, input artifact, output decision, and person-hour budget for every ART event.
Product Management owns backlog and content decisions, Business Owners value PI Objectives, System Team supports integration, and Scrum Masters coach team participation.
Funding proceeds when event overhead is below 8% of ART capacity, every event has a measurable output, and duplicate status meetings are removed within the first PI.

Why interviewers ask this: The interviewer wants an integrated SAFe cadence with explicit costs and accountabilities rather than a calendar full of ceremonies.

Locked questions

21
Eight teams with 74 people use two-week iterations, release monthly, and show a 38-day median feature cycle time on a $11 million annual product group. What cross-team flow dashboard would you create for the next two quarters?
releasesiterationcross-team
22
Six teams with 57 people deploy weekly, plan in two-week Sprints, and have seen product-group work in progress rise from 45 to 78 items over six weeks while throughput remains 18 items per week. How would you use a cumulative flow diagram?
agiledeploymentthroughput
23
Ten teams with 102 people review investments monthly, have 27 initiatives competing for a $12 million annual budget, and can actively support only eight initiatives at once. How would you establish Portfolio Kanban?
agileportfolio
24
Seven teams on one ART plan in two-week iterations, have 14 cross-team dependencies, and need a six-month forecast for a $9 million roadmap in Jira Advanced Plans. How would you configure and govern the plan?
configdependenciesiteration
25
Eight teams with 80 people complete a 10-week PI in two-week iterations, with planned business value of 160 and actual business value of 128 across their PI Objectives. How would you report predictability over the next four PIs?
iteration
26
A nine-team ART leaves PI Planning with 26 program risks, nine without a ROAM state, and stale actions from the previous PI. What operating policy would you establish for program risks?
program-management
27
Six teams with 56 people plan quarterly, have 42 proposed PI Objectives against four company OKRs, and control an $8 million annual roadmap. How would you align work without cascading OKRs into a task tree?
roadmapresilienceroadmapping
28
Ten teams with 98 people select work every 10-week PI, have 18 features competing for a $9 million budget, and can fund only 11 this interval. How would you use WSJF without false precision?
29
Five teams with 46 people deliver an average of 12 similarly sized features per month, plan every two weeks, and need a probability-based forecast for 70 features over the next 16 weeks. How would you build it?
probability
30
Eight teams with 76 people review the portfolio monthly, spend 62% of capacity on features and 8% on technical health, and have a $14 million 12-month plan with feature flow time rising from 30 to 44 days. How would you rebalance economically?
portfoliocapacitycapacity-planning
31
You support seven teams with 65 people and mentor five Scrum Masters every two weeks for six months, while team impediment age ranges from 4 to 18 days. How would you structure the mentoring?
agilementoring
32
Nine teams with 86 people have eight Scrum Masters, work in two-week Sprints, and can invest 90 minutes monthly plus $40,000 annually in a Community of Practice. What would you build?
agile
33
Four Scrum Masters support six teams with 58 people on two-week Sprints, and you have a 12-week coaching window to improve retrospective action completion from 35% to 70%. How would you contract the coaching?
agile
34
Eight teams with 78 people work in two-week Sprints, and leadership wants quarterly maturity evidence over nine months before renewing a $300,000 coaching budget. How would you provide evidence without ranking teams?
agile
35
Six teams with 60 people hold a quarterly 90-minute product-group retrospective, but only 12 people usually speak and cross-team actions close at 40%. Which Liberating Structure would you use over the next two quarters?
agilecross-team
36
Seven Scrum Masters support seven teams with 68 people, meet biweekly, and have three months to improve coaching responses to recurring Sprint Goal changes without adding a $20,000 external course. How would you use Troika Consulting?
agile
37
Five teams with 47 people work in two-week Sprints, answer a monthly safety pulse at only 48%, and have six months to raise speak-up behavior before a $5 million product expansion. How would you improve psychological safety without exposing individuals?
agile
38
Six Scrum Masters support eight teams with 76 people on two-week Sprints, and three new Scrum Masters spend 30% of their time making backlog and assignment decisions during a four-month onboarding period. How would you coach the accountability boundary?
agileonboardingbacklog
39
Four Scrum Masters support four teams with 38 people in two-week Sprints, and you have 12 weeks for a peer-observation program with a $15,000 learning budget. How would you prevent observation from becoming performance appraisal?
agileprogram-managementperformance
40
Nine Scrum Masters support 10 teams with 96 people, meet monthly, and have two quarters to cut Community of Practice backlog age from 110 to 45 days within a $60,000 budget. How would you use Ecocycle Planning?
agilebacklog
41
Eight component teams with 82 people work in two-week iterations, and 72% of product items cross at least three teams, producing a 51-day feature lead time on a nine-month $13 million roadmap. How would you apply Team Topologies?
roadmapcomponentsiteration
42
Six stream-aligned teams with 61 people deploy weekly, each spends 18% of capacity duplicating CI and observability work, and the product group can invest $2 million over 12 months in a platform team. What boundary would you define?
deploymentobservabilitycapacity
43
Seven product teams with 70 people use two-week iterations, face a 16-day wait for cloud-security knowledge, and have 12 months and $1.5 million for migration. Would you create an enabling team?
migrationsiteration
44
Five teams with 49 people release monthly, but six scarce specialists maintain a pricing algorithm used by all teams, and the company has $4 million and 18 months to expand it. Which Team Topologies pattern fits?
pricingalgorithmsreleases
45
Nine teams with 90 people own 31 services, deploy every two weeks, and 28% of changes require three or more approval handoffs over a $15 million annual portfolio. How would you clarify ownership boundaries?
portfoliodeploymentownership
46
Eight teams with 77 people meet weekly in ART Sync, carry 44 organizational impediments with a median age of 21 days, and have two quarters to reach 10 days within a $500,000 improvement budget. What system would you build?
system-design
47
Six teams with 55 people plan monthly, wait a median 17 days for architecture decisions, and have a $9 million 12-month roadmap with two-week delivery iterations. How would you reduce decision latency?
roadmaparchitecturelatency
48
Ten teams with 104 people deliver a regulated banking platform in 10-week PIs and two-week iterations, with $30 million and 18 months available. How would you design governance without adding approval meetings to every Sprint?
agilegovernancedesign
49
Seven teams with 69 people use two-week iterations, have a $1.2 million enablement budget, and need a 12-month roadmap to move from project handoffs toward product ownership without a big-bang reorganization. What would you propose?
roadmapiterationownership
50
Eight teams with 84 people plan every 10 weeks, show 58% of work crossing team boundaries and 12-day median decision latency, and have a $16 million 12-month product-group horizon. How would you combine topology, impediment, and governance changes?
governancelatency
51
Six teams have 18 open dependencies, integrated testing starts in 15 working days, and the release burn-up shows only 62% of forecast scope done. What decision do you drive now?
scope-managementreleasesdependencies
52
Across 5 teams, nine cross-team dependencies have been unresolved for 12 days, two team leads dispute ownership, and a customer release is due in 10 working days. What do you do in the next 24 hours?
releasesdependenciesownership
53
At week 6 of a 10-week PI for 6 teams, 7 of 12 PI Objectives are red after a vendor API slipped by 4 weeks, and Business Owners still expect the original plan. What decision do you facilitate?
procurementapi
54
Four teams need a shared authentication service for a launch in 6 weeks, but the architecture runway is 2 sprints late and integration estimates now exceed the remaining window by 30%. What do you drive?
agilelaunchesauth
55
During day one of PI Planning, eight teams present draft plans with 11 cross-team conflicts, shared specialists loaded at 140%, and two feature-sequencing decisions that breakouts cannot resolve. How do you run management review and problem-solving before day two?
cross-team
56
A strategic customer moves a contractual deadline forward by 3 weeks, 5 teams are already at 88% planned capacity, and missing it carries a $2 million penalty. What is your response?
capacity-planningcapacityestimation
57
The first integrated demo for a 5-team release fails 11 of 20 customer journeys, with 4 weeks left and 27 defects crossing team boundaries. What do you do next?
releasesdefects
58
Ten teams committed to 26 PI Objectives, but capacity data after planning shows 130% load and 40% of key specialists are shared across teams. What decision do you drive before execution starts?
capacity-planningcapacity
59
In a 6-team release, a production certificate dependency owned by another division is discovered 72 hours before go-live, and without it 3 of 7 services cannot start. What do you do?
dependenciesreleases
60
A 7-team program has missed 3 consecutive quarterly releases by 2 to 5 weeks, and its largest customer threatens to leave in 60 days. What intervention do you lead first?
program-managementreleases
61
The CEO asks for a firm 12-month delivery date for a new product after only 2 weeks of discovery, while 6 teams have uncertainty ranges of 40% to 90%. What do you do?
62
A portfolio assigns 14 strategic initiatives to 8 teams for the quarter, but capacity analysis shows 145% demand and only 3 initiatives have validated customer outcomes. What do you facilitate?
portfoliocapacityvalidation
63
Across 6 teams, two value streams both claim first priority on a shared platform team that has 60% of the capacity requested, and the funding committee meets in 48 hours. What do you do?
capacity-planningcapacity
64
A VP publishes a monthly ranking of 9 teams by velocity, and the bottom 3 teams have started inflating estimates by about 35%. What is your response?
delivery-metricsestimation
65
Leadership plans to tie 20% of annual bonuses for 7 teams to story-point growth, and the policy launches in 3 weeks. What do you do?
launches
66
Halfway through a fixed annual plan for 6 teams, customer data shows that 4 of 9 funded features have under 5% adoption, but executives insist all 9 must ship by year-end. What do you facilitate?
decision-making
67
Three weeks before a board review, the probability of delivering a named release scope by its date falls from 80% to 45% under the current assumptions, and $6 million of expected revenue depends on it. How do you communicate the risk?
risk-managementscope-managementreleases
68
In an 8-team portfolio, a director asks you to remove 6 red risks from the dashboard before an investor update in 2 days, although 4 have no mitigation owner. What do you do?
portfoliorisk-management
69
A six-team ART depends on three database specialists in Shared Services who are available at 40% capacity, two specialists leave after this PI, and nine Features need their expertise. How would you plan capacity and knowledge transfer?
capacity-planningdatabasecapacity
70
A new regulation consumes an estimated 25% of capacity across 8 teams, but portfolio leadership refuses to alter 17 quarterly commitments due in 9 weeks. What do you facilitate?
portfolioestimationcapacity
71
An 8-team ART completes every SAFe ceremony, yet only 41% of PI Objectives are achieved and decision lead time has risen from 4 to 11 days over 2 PIs. What do you change first?
72
Six teams release monthly, deployment takes nine days, manual integration and approvals create 70% of the delay, and recovery from a failed release averages six hours. How would you use CALMR to choose the first Continuous Delivery Pipeline bottleneck?
deploymentci-cdtracking
73
Six component teams hand work through 4 queues, median end-to-end lead time is 47 days, and waiting accounts for 31 days. What intervention do you facilitate?
componentse2edata-structures
74
A weekly 90-minute Scrum of Scrums for 10 teams has become a status round, while 13 dependencies are older than 2 weeks. What do you do next week?
agiledependencies
75
All 7 teams report over 90% sprint-goal success, but system lead time has doubled to 36 days and integrated defect escape has reached 12%. What do you address?
agilesystem-designdefects
76
Five teams on one ART report 95% predictability, but an audit finds they remove unfinished scope one day before every review and reclassify 18% of objectives after the fact. What do you do?
scope-management
77
Business Owners give every objective a score of 10 after 4 teams delivered only 6 of 11 intended outcomes, making the PI report look perfect. What do you do?
78
On a nine-team ART, the System Team requires every team to hand off branches during the final iteration, so an integrated Increment appears only every 10 weeks and 31 integration defects remain open. How would you redesign the System Team role?
system-designiterationdefects
79
Lean Portfolio Management asks for a SAFe Measure and Grow scorecard across an eight-team ART, and the PMO proposes aggregated story points as the primary measure. What would you include instead?
portfolioaggregation
80
A $4 million portfolio Epic has an executive sponsor but only untested customer and revenue assumptions, and leadership wants all 10 teams to begin implementation next quarter. What do you facilitate before full funding?
portfoliosponsor
81
After a 6-month agile transformation across 8 teams, lead time remains at 52 days, engagement fell 9 points, and leaders say teams are resisting change. What do you do first?
agileengagement
82
Three directors cancel retrospectives for 6 teams after two sprints, saying the 6 hours per team produced no visible value, while repeated impediments rose from 4 to 11. How do you respond?
agile
83
A 600-person division has ordered a SAFe rollout, trained 30 change agents, has not identified development value streams, and wants the first ART launched in 12 weeks. How would you use a LACE and the SAFe Implementation Roadmap?
launchesreleasesroadmap
84
Across 8 teams, 3 Scrum Masters resolve most impediments themselves, 2 act mainly as project trackers, facilitation satisfaction ranges from 38% to 91%, and 14 program impediments are overdue. What do you do in 60 days?
agileprogram-management
85
An eight-team ART uses every Innovation and Planning Iteration to finish integration, regression testing, and 38 carried defects, leaving no time for learning or innovation. How would you restore its purpose?
regressiondefectsiteration
86
Five teams span 7 time zones, yet 80% of cross-team decisions happen in one 6 p.m. Europe meeting that excludes 14 of 38 people. What do you do?
cross-team
87
In PI Planning for 6 teams across 4 countries, one location speaks for 72% of plenary time and the other locations raise 9 critical risks only after the confidence vote. What do you change?
88
Four middle managers assign daily tasks directly to 5 teams, unplanned work is 33%, and team-owned sprint goals have failed for 4 consecutive sprints. What do you do?
agile
89
After 2 failed reorganizations in 18 months, 7 teams refuse a new value-stream pilot and only 19% believe leadership will keep its commitments. How do you proceed?
90
A 3-team transformation pilot cut lead time by 28%, but 5 non-pilot teams report 11 unresolved constraints that the pilot avoided, and the sponsor wants rollout next Monday. What do you do?
releasessponsor
91
A 96-minute outage spans 7 teams, causes $1.8 million in lost transactions, and the first review draft blames one team's deployment. How do you facilitate learning?
transactionsdeployment
92
Four similar incidents occurred across 6 teams in 90 days, and 17 of 24 retrospective actions remain open beyond their due dates. What do you do?
agileincidents
93
A $40 million portfolio funds three value streams through annual project budgets, priorities change quarterly, and Finance wants to introduce Lean Budgets. How would you establish guardrails and participatory budgeting without displacing investment authority?
guardrailsbudgetportfolio
94
A critical zero-day affects 24 services owned by 8 teams, exploitation is confirmed in the industry, and Security requires mitigation within 48 hours. What do you do?
risk-management
95
An audit in 30 days finds that 5 of 9 teams cannot trace 38 production changes to approvals, although the software itself is compliant. What do you do?
96
A director asks 6 teams to exclude blocked days from lead-time reports so a promised 20% improvement appears achieved before the board meeting in 4 days. What do you do?
promises
97
The agile management tool used by 10 teams is unavailable for 48 hours during PI execution, and 63 active dependencies plus 120 work items are inaccessible. What do you do?
agiledependencies
98
A sole-source vendor misses three integration milestones by 12 days, blocking four teams and leaving a 55% probability of delivering the named $4 million launch scope by its date under current assumptions. What do you facilitate?
milestonesprocurementplanning
99
A forced migration to a new planning tool starts in 2 weeks for 10 teams, but the pilot lost 14% of dependency links and doubled update time from 4 to 8 minutes. What do you do?
dependenciesmigrations
100
With 18 days left, an 8-team release faces a critical security fix on 9 services, a vendor delay of 7 days, 6 red PI Objectives, and a $3 million customer deadline, while confidence is 2.6 out of 5. What do you do in the first 24 hours?
procurementreleasesestimation

Questions

1A $24 million product group has 10 teams and 96 people working on three products over a shared platform, uses two-week iterations, and funds work quarterly for the next 18 months. Which scaling approach would you choose without importing an entire framework by default?

2Six feature teams with 54 people build one commerce product from one Product Backlog, run two-week Sprints, and must cut feature lead time from 42 to 25 days within nine months on a $7 million roadmap. Would you use LeSS, SAFe, Nexus, or lightweight coordination?

3Five Scrum Teams with 45 people deliver one identity platform in two-week Sprints, but 40 cross-team integration defects appear each quarter and the next product release is six months away. Which scaling framework would you introduce and what would it cost?

4Four autonomous product teams with 36 people deploy weekly, share only two dependencies per month, plan in two-week Sprints, and have a six-month $3 million growth horizon. Would you add a scaling framework?

5Eight teams with 78 people are building one regulated medical product on two-week iterations, face quarterly evidence reviews, and have 12 months and $18 million to reach approval. How would you adapt SAFe without turning compliance into extra ceremony?

6An eight-team ART depends on four security and data specialists who can each give 30% to 50% of their time, and it also consumes a deployment platform run by a dedicated product team. How would you represent these two kinds of support in PI Planning?

7Ten feature teams with 92 people work on one global marketplace, deploy weekly, use two-week Sprints, and must halve a 46-day idea-to-market time within 12 months on a $16 million budget. Would you choose LeSS Huge or Essential SAFe?

8Nine teams with 84 people combine hardware, firmware, and cloud work, integrate every 12 weeks, use two-week team iterations, and have a $25 million 18-month product horizon. Which coordination model would you select?

10Three ARTs with 24 teams and an external signaling supplier must deliver one integrated rail-control solution every six months, and no ART can validate the customer outcome alone. When is a Solution Train actually justified?

11Eight teams with 76 people across two time zones are entering a 10-week Planning Interval, work in two-week iterations, and must deliver a $6 million customer outcome in six months. How would you design PI Planning?

12Six teams with 54 people plan an eight-week PI in two-week iterations, have 30 candidate features, and need four measurable product outcomes within a $5 million half-year roadmap. How would you produce useful PI Objectives?

13Nine teams with 86 people use two-week iterations inside a 10-week PI, reserve between 5% and 30% for support, and have a $9 million release due in five months. How would you make capacity planning comparable without standardizing velocity?

14Seven teams with 64 people have 48 planned dependencies across two-week iterations and a five-month release horizon, with current dependency age at 18 days. How would you build and operate the PI dependency board?

15Eight teams with 80 people finish two-day PI Planning for a 10-week interval, but the first confidence vote averages 2.8 out of 5 and the roadmap commits $7 million over six months. What should happen before planning closes?

16Ten teams with 100 people work in two-week iterations, carry 35 active dependencies, meet on a 10-week PI cadence, and have a nine-month delivery horizon. How would you design ART Sync?

17Seven teams with 68 people build an integrated service in two-week iterations, plan every 10 weeks, and have a $10 million release in eight months. How would you design the System Demo?

18Nine teams with 90 people close a 10-week PI after five two-week iterations, have a 24-day median feature lead time, and need to reach 16 days within 12 months. How would you run Inspect and Adapt?

19Six teams with 58 people span five time zones, use two-week iterations and a 10-week PI, and have a $400,000 annual coordination budget. How would you make distributed PI Planning inclusive and affordable?

20Eight teams with 75 people use two-week iterations, a 10-week PI, and a six-month $8 million roadmap, but leadership wants one coherent ART event calendar before funding. What cadence would you propose?

Locked questions