Infrastructure
110 bites tagged Infrastructure — interview questions with model answers, and 60-second explainers.
What is IaC and its CI/CD benefits over manual provisioning?
Cover descriptive models, idempotency, and on-demand environments for repeatable CI/CD. Your understanding of IaC as versioned, declarative infrastructure that stops snowflake environments and drift.
Explain blue-green deployment, its prerequisites, and how it reduces deployment risk.
Tests operational maturity for zero-downtime cutover. Great answers: parallel environments, load balancer switching, backward-compatible schemas, externalized state, and instant rollback vs partial in-place failure.
What is Infrastructure as Code (IaC), and how does it support CI/CD?
This tests if you link declarative definitions to repeatable pipelines. A strong answer covers idempotence, versioned templates, and preventing snowflake environments. A red flag is calling IaC mere scripting without CI/CD integration.
How would you structure a backend architecture A/B test and define metrics?
This tests causal inference rigor for infrastructure changes. A strong answer covers sticky user routing, controlling for geography and time, and paired primary metrics like P99 latency and error rate.
Vercel ships AI Gateway, Workflow SDK, and Sandbox for agents
Vercel Agent Stack unifies hundreds of models, durable workflows, and isolated microVMs for production agents. One endpoint routes across providers, checkpoints resume failed runs, and sandboxes isolate unreviewed code.
Vercel Ship 2026: agent stack, eve framework, and microservices
Vercel Ship 2026 unveils open-source eve framework and Vercel Services for microservices July 1. Vercel is repositioning as agentic infrastructure with Connect credentials and expanded Python backends. Evaluate if your next agent deploys here instead of AWS.
Next.js 16.2 ships stable Adapter API for all platforms
Next.js 16.2 ships a stable Adapter API co-built with OpenNext, Netlify, and Cloudflare. The typed build contract lets any platform target full framework fidelity using the same public hooks Vercel uses. Stop reverse-engineering build output.
Design a cost-aware ML training platform for heterogeneous hardware
Tests hardware abstraction and cost-aware cross-accelerator scheduling. Strong answers cover a device-agnostic spec, a performance predictor, a cost-per-step model, and bin-packing against spot prices. Red flag: ignoring per-step cost and migration overhead.
Design a centralized model registry for a large enterprise
Tests ML artifact governance at scale. Strong answers cover immutable versioned artifacts with dependency manifests, a framework-agnostic API, and pluggable deployment targets. Red flag: treating models as opaque files without environment reproducibility.
Describe a Model Registry and how it differs from versioned storage
It tests governance and lifecycle metadata beyond file storage. A strong answer covers lineage, stage transitions, approval gates, and artifact metadata, contrasting with buckets that only store file versions.
Describe an ML workflow with massive egress fees and re-architecture to mitigate
Tests whether you recognize egress spikes when storage and compute cross cloud or region boundaries. Great answers sketch a multi-cloud training pipeline, cite per-GB rates, and propose caching or compute placement. Red flag: suggesting compression alone.
Design a showback or chargeback system for ML infrastructure costs
Tag workloads to cost centers; define shared-resource formulas; automate reconciliation; use showback. Bridging ML telemetry with FinOps for shared GPU storage. Using raw cloud bills as attribution without GL mapping.
Describe a basic lifecycle policy to manage cloud storage costs
This tests cost optimization via tiered storage and automated expiration. Strong answers list transitions from Standard to IA to Glacier, then deletion after set days, plus retrieval costs. A red flag is using manual scripts instead of native lifecycle rules.
Design a defense-in-depth strategy against adversarial evasion on a deployed image classifier
Proactive: adversarial training, preprocessing, ensembles. Your ability to layer training-time and inference-time defenses for adversarial robustness.
What trade-offs decide managed ML platforms versus open-source Kubernetes?
Weigh total cost plus hidden engineering headcount, lock-in vs flexibility, and audit feature gaps. Ops overhead vs speed for ML infra. Recommending open-source purely to cut cost while ignoring the 2-4 person tax.
What is a feature store and how does it prevent training-serving skew?
This tests training-serving consistency via centralized feature management. Covers offline batch storage, online serving, shared transformations, and alternatives like ad-hoc ETL. A red flag is calling it just a database and ignoring point-in-time correctness.
What are the essential components of an end-to-end ML platform?
Tests mapping the ML lifecycle to infrastructure. A strong answer walks through six stages: data ingestion, feature store, training pipeline, model registry, deployment/serving, and monitoring with feedback loops.
How would you architect dynamic training resource provisioning?
Routes jobs to right-sized instances, provisions managed jobs using spot, and auto-tears down on completion. elastic ML infrastructure and cost control. suggesting always-on clusters or serverless GPU training.
How do you monitor thousands of per-customer models as a fleet?
Tests fleet-level statistical aggregation versus per-instance alerting. Strong answers propose tiered telemetry, cohort baselining for drift, and hierarchical alerting to prevent fatigue.
Differences between monitoring a traditional REST API and a production ML model
Contrast latency/errors with ML signals like data drift and training-serving skew against baselines, noting ground truth delays. Awareness that ML fails via data decay, not code bugs.
Design a multi-model serving architecture for infrequently used models
Tests sparse-traffic cost efficiency via shared compute and dynamic loading. Strong answers: LRU cache on fast storage, scale-to-zero with async preload, pinned hot models, per-model quotas.
Design cost-effective inference for spiky traffic without idle GPUs
Tests designing inference that cuts idle GPU cost during troughs yet handles spiky peaks with low latency via SageMaker blue/green fleets, production variants, and CloudWatch baking periods. Red flag: always-on GPU pools with naive auto-scaling.
How would you design an A/B test for two live ML models?
Tests production experimentation rigor beyond random splitting. Strong answers cover: consistent user hashing for sticky assignment, isolated feature stores, guardrail metrics, and pre-calculated statistical power.
How would you systematically diagnose high latency in an online inference service?
Check p90/p99 and TTFT to split queuing from compute; inspect queue depth, batch size, GPU, and benchmarks; check cache. Systems reasoning across serving stack.
Get Infrastructure bites daily.
Five a day, five minutes, offline. With quizzes so it sticks.
Open testing — you’ll join as an early tester.