yoklainterview sim
ProgressHistory

Data Engineer Senior Interview Questions

392 verified Data Engineer Senior interview questions — solve with answers, learn from explanations, test yourself in a real simulation.

Try the real simulation →

Sample questions

Data Governance LineageDifficulty 3
How does 'data ownership' typically differ from 'data stewardship'?
  • aOwnership is a purely technical role while stewardship is purely legal
  • bThey are two names for exactly the same responsibility
  • cThe owner is accountable overall; a steward handles day-to-day tasks✓
  • dStewardship only applies to datasets that contain PII
Explanation:Ownership carries overall accountability for a dataset, while stewardship often covers the operational, day-to-day quality and definition work, frequently on the owner's behalf.
Data Governance LineageDifficulty 3
A company deletes raw event data after 90 days per policy, but later discovers it needs that data to investigate a billing dispute from four months ago. What does this reveal?
  • aThe lineage graph for the event data was incomplete
  • bThe data catalog entry for the event data was missing a description
  • cThe access control rules for the event data were too permissive
  • dThe retention policy lacked an exception for legitimate longer-retention needs✓
Explanation:This shows the retention policy failed to account for a legitimate need to retain data longer for dispute investigation, or lacked an exception process.
Data Governance LineageDifficulty 3
A table's schema changes from
before: order_id, amount
after: order_id, amount_cents

Several downstream reports still reference amount. Without a downstream impact analysis before the change, what is the most likely outcome?
  • aThe reports will automatically convert amount_cents back to amount
  • bThe downstream reports will break or silently show wrong values✓
  • cThe change will have no effect since column renames are always backward compatible
  • dThe data catalog will block the schema change automatically
Explanation:Without checking downstream impact first, reports that still reference the old column will break or silently show incorrect values.
Data Governance LineageDifficulty 3
What is a common drawback of a purely centralized data governance model at a large organization?
  • aThe central team can become a bottleneck for fast decisions✓
  • bIt guarantees every team will duplicate the same datasets
  • cIt removes the need for any data ownership at all
  • dIt automatically eliminates the need for a data catalog
Explanation:A common downside of pure centralization is that the central team becomes a bottleneck for domain teams needing fast, context-specific decisions.
Data Governance LineageDifficulty 3
A large company has many independent domain teams, each best positioned to know their own data's quality rules, but the company still wants consistent classification and access standards company-wide. Which governance approach best fits this need?
  • aA fully centralized model where one team manually reviews every dataset
  • bNo governance model at all, letting each team improvise
  • cA model where governance is only applied to the largest datasets
  • dA federated model: central standards, domain teams own their own data✓
Explanation:A federated model combines company-wide standards and shared tooling with domain teams owning their own data within that framework, balancing consistency and speed.
Data Modeling WarehousingDifficulty 3
A retail fact table is designed at 'one row per store per day' grain. Finance later asks for analysis of individual, transaction-level fraud patterns. Which statement is correct?
  • aThe existing grain is sufficient, daily totals hold enough detail
  • bGrain can be changed later with no impact, by adding columns
  • cA new fact table at transaction grain is needed for this✓
  • dFraud detection should be done exclusively inside the dimension tables
Explanation:Once data has been aggregated to a coarser grain, the finer-grained detail is gone; a transaction-level analysis requires a fact table built at transaction grain from the start.

Topics

Data Governance Lineage 75Data Modeling Warehousing 75Data Partitioning Scaling 75Data Pipeline Design 75Data Pipeline Orchestration 75Data Pipeline Reliability 75Data Quality Validation 75Streaming Fundamentals 75

By seniority

Data Engineer JuniorData Engineer MidData Engineer Senior

Other fields

BackendFrontendFullstackDevOps / CloudAI EngineerDatabaseMobileSecurityQA / Test AutomationGame DevelopmentML Engineer

Test yourself against the 600-question Data Engineer bank.

Start interview