You are asked to lead the migration of 200 services to the cloud. What is your strategy and how do you de-risk it?
What they are really testing: Whether you think like a principal: strategy, sequencing, risk, and people, not a server checklist. They want a phased, evidence-driven plan and how you carry an organization, not just workloads.
A real interview question
You are asked to lead the migration of 200 services to the cloud. What is your strategy and how do you de-risk it?
What most people say
drag me
“I would lift-and-shift everything to the cloud as fast as possible to get it done.”
Lift-and-shift-everything ignores that services differ, skips the landing-zone foundation, has no risk strategy or rollback, and forgets the people. It optimizes for "done" over "succeeded", the opposite of principal judgment.
The follow-ups they ask next
A critical legacy service is too risky and expensive to move. What do you do?
Retain or retire is a valid principal answer, not everything must move. Justify with cost/risk, possibly wrap it (anti-corruption layer) or replatform later; don’t force a migration that destroys value to satisfy a mandate.
How do you keep 200 teams from each building cloud their own way?
Golden paths / paved roads: opinionated, self-service platform templates, landing zones, and guardrails (policy-as-code) so the easy way is the compliant way. Enablement and a platform team, not a 200-page standards doc nobody reads.
What the interviewer is listening for
- Segments the portfolio
- Uses the 6 R’s deliberately (incl. retire/retain)
- Foundation + golden paths first
- Incremental, reversible, evidence-gated
- Thinks about people and cost
What sinks the answer
- "Lift-and-shift everything fast"
- No landing zone / foundation
- No rollback or wave gating
- Forgets upskilling and org change
- Treats all 200 identically
If you genuinely do not know
Say this instead of freezing. Reasoning out loud from what you do know beats silence every single time, and a good interviewer is listening for exactly that.
“I’d [segment the 200 by dependency/criticality/effort], assign [a 6-R pattern per service, including retire/retain], lay [the landing zone + golden paths first], then sequence [low-risk waves first to prove the path], each [incremental, reversible, gated on success not a date], while [upskilling teams and tracking cost/reliability].”
Keep going with strategy & leadership
Principal
A few teams want an internal feature-flag and config service. You could build one on top of your cloud primitives or buy a SaaS. How do you make the build-vs-buy call for a platform capability the whole org will depend on?
Principal
You have 30 teams each spinning up cloud infrastructure their own way, with wildly different quality. How do you establish golden paths or paved roads so teams stop reinventing the wheel without grinding delivery to a halt?
Principal
Cloud spend is growing faster than revenue and leadership wants it under control across dozens of teams. How do you set up org-wide cost governance, or FinOps, without turning into the team that says no to everything?
Principal
Design a globally distributed, multi-region active-active system. Walk me through the hard parts.
Principal
The org is moving to many cloud accounts across many teams. How do you design the multi-account landing zone and guardrails so teams move fast safely, instead of locking everything down or letting it become the wild west?
Principal
Design a global rate limiter that protects your APIs across many regions, enforcing per-customer quotas. Walk me through the design and, more importantly, the trade-offs you would make and how you would sequence building it.
Knowing the answer is not the same as recalling it under pressure
Sign in to send the questions you fumble to spaced recall, so they come back right before you would forget them, and learn the concepts behind them with hands-on labs.
Start free