Evidence-based ML interview readiness check

No account, dates, identity, or answer text. Selections and calculations stay in this browser. Only coarse “started” and “completed” event names are counted.

1. Target and capacity

Role Target level Primary domain Hours available each week Weeks until the loop

2. Actual interview rounds

Select the formats the recruiter described, not every topic you might study. Domains such as LLMs and recommendations belong above, not here.

ML breadthMechanisms, model choices, metrics, and “why” follow-ups.ML implementation / codingML primitives, evaluation systems, debugging, and resource-aware implementation under time pressure.ML system designRequirements, data, modeling, evaluation, serving, monitoring, and iteration.Project deep-diveOwnership, technical decisions, failures, evidence, influence, and reflection.Behavioral / leadershipConflict, prioritization, failure, influence, mentoring, and collaboration.Product & experimentationMetrics, experiment design, causal threats, guardrails, and ship decisions.Research depth / paper critiqueHypotheses, ablations, evidence quality, derivations, and research judgment.Systems / infrastructureTraining and serving scale, reliability, bottlenecks, and cost trade-offs.

3. Current evidence

Rate the best evidence you have from a recent timed, closed-book attempt. “I understand it when reading” is not a 3.

ML fundamentalsProduction & debuggingLLM systemsRetrieval, search & recommendationsML system designProduct & experimentationBehavioral & leadershipMath & research depthML implementation

Estimate the work from evidence, not confidence.

1. Target and capacity

Recommended plan

Your first three gaps

Do not graduate a critical gap on confidence alone