Pick a focused question that fits your time, stack, and interview goal.
How much time do you have?
Show one-drill sessions you can finish now.
1225 drills fit a 20-minute session. Closest fits first.
Page 6 of 52
Carries source permissions into ingestion and query-time filtering while failing closed across retrieval, caching, and citations.
Evaluates verified outcomes, trajectory quality, safety, efficiency, recovery, and important slices in realistic environments.
Separates storage from processing location, verifies provider and feature-specific retention, and minimizes sensitive inputs.
Judges claims against allowed evidence, separates correctness from support, and tests whether the system answers or abstains appropriately.
Builds a query-and-relevance dataset and separates retrieval recall, ranking quality, context quality, and generated-answer quality.
Measures schema validity, tool selection, argument correctness, policy, execution outcome, repair, and final user result.
Controls admission, concurrency, queues, tenant fairness, token budgets, and overload behavior before provider throttling cascades.
Maps DRF and application failures into stable client errors while preserving internal diagnostic evidence.
Makes nested create and update semantics, authorization, identity, and transaction ownership explicit.
Uses FastAPI Security dependencies to combine declared scopes with current application authorization.
Separates fast rollback isolation from tests that must observe commits, multiple connections, or database concurrency.
Combines application authorization, scoped platform resources, tenant budgets, admission control, usage attribution, and noisy-neighbor protection.
Uses representative traffic and tail latency to find the real bottleneck and a safe operating range.
Uses stable operation identities, persisted states, deduplication, side-effect keys, and replay rules across requests and workers.
Rebuilds and compares incompatible vector spaces using versioned indexes, shadow queries, atomic cutover, and rollback.
Uses supported version steps, ng update migrations, dependency compatibility, targeted cleanup, and staged production evidence.
Constrains source-to-sink combinations with least privilege, egress policy, argument checks, approval, and information-flow awareness.
Carries trusted tenant scope through retrieval, caches, memory, tools, traces, evaluations, and asynchronous execution.
Chooses evidence based on whether latency comes from CPU execution, waiting, contention, or an external dependency.
Uses infrastructure and configuration as code, immutable behavior versions, environment-specific resources, evaluation gates, and drift detection.
Checkpoints authoritative state, reconciles ambiguous tool outcomes, preserves approvals, and resumes without repeating completed work.
Builds system-specific adversarial testing around realistic assets, identities, channels, tools, and measurable impact.
Contains compromised AI capabilities, preserves evidence, scopes derived data and actions, remediates boundaries, and verifies safe recovery.
Uses explicit capability and risk policy to route by task while preserving evaluation, version attribution, budgets, and fallback semantics.