Design a tiered API rate limiter
WHAT IT TESTS: rate-limiting algorithms and distributed counting. OUTLINE: pick token bucket or sliding-window, key limits by partner tier, track counters in a shared store like Redis, decide at the edge. RED FLAG: per-node counters bypassed across instances.
Public API design versus internal API design
WHAT IT TESTS: public APIs are long-lived contracts. OUTLINE: public needs strict versioning, long deprecation, scoped auth like OAuth and API keys, and polished docs; internal can move faster.
Decommission a depended-on API gracefully
WHAT IT TESTS: graceful API deprecation with live consumers. OUTLINE: map consumers and usage, provide a migration target, announce a versioned timeline, decommission only after traffic hits zero. RED FLAG: shutting off the API with no replacement.
Unify behavior, billing, and CRM data
WHAT IT TESTS: an ELT pipeline with identity resolution. OUTLINE: ingest events, sync Stripe and Salesforce into a warehouse, resolve identities to one customer, model unified metrics. RED FLAG: ignoring identity stitching and the anonymous-to-known join.
Prepare for a launch traffic spike
WHAT IT TESTS: capacity planning for a known surge. OUTLINE: model expected load, load-test to find the first bottleneck, scale and cache, add graceful degradation and a queue for spiky writes. RED FLAG: turning up autoscaling without finding the bottleneck.
Instrument an onboarding flow for analytics
WHAT IT TESTS: funnel instrumentation and event design. OUTLINE: track each onboarding step plus the activation milestone, define a clean event schema with stable IDs, send reliably via batching or server-side. RED FLAG: logging vanity counts with no funnel.
Evaluate a high-risk full rewrite
WHAT IT TESTS: judgment on big-bang rewrites. OUTLINE: surface rewrite risk with POCs, propose incremental displacement like Strangler Fig, quantify the cost of stalled features. RED FLAG: endorsing an 18-month freeze with no phased alternative.
Build the case to deprecate a legacy feature
WHAT IT TESTS: data-driven deprecation with empathy. OUTLINE: quantify cost versus value and who the 2% are, propose migration paths and a phased sunset, weigh velocity against trust. RED FLAG: killing it abruptly on raw usage with no migration.
Estimate and de-risk an ambiguous initiative
WHAT IT TESTS: estimating under uncertainty. OUTLINE: decompose into phases that front-load learning, run spikes to retire risk, communicate estimates as ranges tied to milestones. RED FLAG: giving a single precise date for work you do not yet understand.
Add a Pro plan and gate features
WHAT IT TESTS: clean feature gating and entitlements. OUTLINE: model plans and entitlements as data, enforce server-side via a central check, decouple the gate from feature code. RED FLAG: scattering isPro checks through the codebase or trusting the client.
Architect today for a loosely defined future
WHAT IT TESTS: evolvable architecture without over-engineering. OUTLINE: isolate volatility behind stable interfaces, use ports and adapters, keep changes reversible and deferred.
Product strategy versus go-to-market strategy
WHAT IT TESTS: what to build versus how to sell it. OUTLINE: product strategy defines the product and roadmap; GTM defines launch, pricing, channels, and audience; they overlap at positioning.
Resolve a low-code versus custom-build conflict
WHAT IT TESTS: principal judgment under a strategic disagreement. OUTLINE: build a spike testing real constraints, surface lock-in and exit cost, propose a hybrid scoped by differentiation. RED FLAG: picking a side ideologically instead of de-risking.
Quantify tech debt and pitch it to a PM
WHAT IT TESTS: translating engineering risk into business language. OUTLINE: quantify probability times impact, tie debt to velocity or incident cost, and propose scoped phased work. RED FLAG: demanding a rewrite on faith using only the words it is messy.
Platform team metrics versus product team metrics
WHAT IT TESTS: platforms serve developers, not end users. OUTLINE: measure adoption, reliability SLOs, integration time, and self-service ratio over user engagement. RED FLAG: copying consumer metrics like DAU onto a platform with no end users.
Shared component library versus per-product builds
WHAT IT TESTS: reuse versus autonomy. OUTLINE: shared libraries cut duplication and enforce consistency but add coupling, versioning, and a coordination tax; per-product code is fast but drifts.
Move from flat to usage-based billing
WHAT IT TESTS: accurate, scalable metering and billing. OUTLINE: reliable usage capture, idempotent aggregation into billing periods, and reconciliation with the provider. RED FLAG: tallying charges synchronously or ignoring duplicates and late events.
Measure a competitor's public performance
WHAT IT TESTS: black-box web performance measurement. OUTLINE: synthetic audits via Lighthouse and WebPageTest, timed public-API probes, and reading response headers. RED FLAG: proposing scraping or load attacks that break terms of service.
Engineering input in a Jobs to be Done workshop
WHAT IT TESTS: thinking in user jobs and outcomes, not features. OUTLINE: frame the underlying job and measurable outcomes the user wants, decouple from any solution, then let features compete to serve them.
Technically analyzing a competitor's product
WHAT IT TESTS: turning competitive technical analysis into roadmap decisions. OUTLINE: probe their stack, performance, APIs, and architecture via public signals and ethical inspection, identify gaps and parity needs, feed differentiation and risk into the…