96 ETL Data Quality and Validation Interview Questions

All 96 questions

How do you ensure data quality and reconciliation when integrating multiple sources into a reporting pipeline?
Asked in 4 of 307 interviews · last asked Oct 23 · Scenario · Mid-level
Show answer
I define shared data rules, standardize source formats, and validate data at ingestion, after transformation, and before reporting. I check required fields, duplicates, valid values, and links between records. I reconcile row counts and business totals against each source, allowing for expected transformation changes. I quarantine bad records and block publication when critical checks fail.
- Agree on shared keys, field meanings, and validation rules with data owners.
- Compare counts and totals for the same time window, accounting for filters, duplicates, and late arrivals.
- Keep source identifiers and audit logs so differences can be traced.
- Automate checks and alerts; assign owners to investigate failures.
What the interviewer checks: The interviewer is checking whether you can detect bad data, explain reconciliation differences, and prevent unreliable reports.
Other ways it was asked
- “How do you ensure the quality of data and the pipeline?”
- “How do you ensure data quality throughout an ETL pipeline?”
- “How do you ensure data quality and reconciliation across a data pipeline? Give an example of how you have done this.”
How would you handle missing fields in data scraped from a website?
Asked in 2 of 307 interviews · last asked Aug 29 · Scenario · Mid-level
Other ways it was asked
- “How do you handle missing or inconsistently formatted fields in scraped data?”
When importing database rows into an application, how would you generate a unique key and validate missing values?
Asked once · last asked Aug 22 · Scenario · Mid-level
Other ways it was asked
- “When importing database rows into an application, how would you generate a unique key and assign values for missing fields?”
How would you address data integration and data quality risks in a financial dashboard?
Asked once · last asked Feb 18 · Scenario · Senior
How do you design data-quality or integrity tests for data pipelines?
Asked once · last asked Mar 3 · Scenario · Mid-level
Other ways it was asked
- “How do you distinguish functional tests from data-quality tests in a data pipeline?”
How would you validate that all rows from a source have arrived in a data warehouse?
Asked once · last asked Mar 3 · Scenario · Mid-level
How should a data pipeline handle a schema mismatch: continue processing with the existing schema, or stop and report an error?
Asked once · last asked Apr 14 · Scenario · Mid-level
When extracting and transforming data for AI model training, what are your key considerations for data quality, consistency, and security?
Asked once · last asked Apr 17 · Scenario · Mid-level
When you validate incoming data in Python, what checks do you treat as hard failures versus quarantined bad records, and how do you route or report them?
Asked once · last asked Apr 25 · Scenario · Senior
In Python, how would you implement a reusable data quality framework with rules, severities, and metrics that works across multiple pipelines while still supporting dataset-specific checks?
Asked once · last asked Apr 25 · Design · Senior
How would you validate an ETL pipeline end to end without relying on developers?
Asked once · last asked May 3 · Scenario · Mid-level
In production, how do you handle schema evolution, such as added or renamed fields and type changes, across ingestion, transformation, and downstream consumers without breaking the pipeline?
Asked once · last asked May 3 · Scenario · Senior
If parent and child records can arrive independently, how would you build a database framework or report to identify child records that have no corresponding parent?
Asked once · last asked May 19 · Scenario · Mid-level
How would you calculate checksums during a data transfer, and what would you do if they did not match because a file was corrupted?
Asked once · last asked May 21 · Scenario · Mid-level
How would you handle integration errors and duplicate data?
Asked once · last asked Jun 21 · Scenario · Mid-level
How would you monitor data quality in pipelines, and what checks or alerts would you set up to catch errors or anomalies before data reaches end users?
Asked once · last asked Jun 24 · Scenario · Mid-level
Other ways it was asked
- “How did you monitor data quality in your pipelines, and what checks or alerts did you set up to catch errors or anomalies before data reached end users?”
How would you address data quality issues and pipeline failures?
Asked once · last asked Jul 16 · Behavioral · Senior
Other ways it was asked
- “How have you addressed data quality issues and pipeline failures?”
How do you handle schema changes when consuming CDC events into a lake destination?
Asked once · last asked Jul 16 · Scenario · Senior
How would you handle added, removed, or renamed columns when applying CDC events to Delta Lake?
Asked once · last asked Jul 16 · Scenario · Senior
How would you ensure data integrity and engineer robust features when integrating data from sources with varying data quality and schema inconsistencies for a predictive maintenance model?
Asked once · last asked Jul 16 · Scenario · Senior
How do you validate the accuracy of extracted content when passing PDF text to AI, and why might you choose not to use PDF-to-text libraries or IBM Docling?
Asked once · last asked Jul 17 · Scenario · Senior
What tricky edge cases can arise with type conversions or schema checks in a data pipeline, and how would you handle them?
Asked once · last asked Jul 21 · Scenario · Mid-level
How does an end-to-end sales validation workflow process incoming records, apply validation logic, and produce its outputs?
Asked once · last asked Jul 27 · Design · Senior
How do you validate that trading data maps accurately from a trading system to regulatory reporting agents?
Asked once · last asked Jul 28 · Scenario · Mid-level
When you find anomalies caused by upstream systems rather than your pipeline, how do you decide whether to correct them in your layer or escalate them, and how do you prevent silent fixes from masking source issues?
Asked once · last asked Jul 29 · Scenario · Senior
How do you design data quality gates, thresholds, and pass/fail criteria so a refresh can automatically block or alert when anomalies spike?
Asked once · last asked Aug 1 · Scenario · Senior
When you receive a refreshed extract each day, how would you set up data quality checks that detect schema drift, missing or renamed columns, type changes, and fail the pipeline with clear diagnostics?
Asked once · last asked Aug 5 · Scenario · Senior
How do you ensure data consistency and correctness across the bronze, silver, and gold layers, especially when source schemas change frequently or source files are corrupted?
Asked once · last asked Aug 10 · Scenario · Senior
How would you handle frequent schema evolution in Azure Synapse Analytics without breaking existing queries or pipelines, and what features or techniques would you use to ensure seamless schema evolution?
Asked once · last asked Aug 11 · Scenario · Senior
In an ML data pipeline, what concrete checks would you add to detect training-serving skew early, and where would you enforce them in code?
Asked once · last asked Aug 13 · Scenario · Senior
When curating large volumes of research artifacts, such as PDF tables and analysis spreadsheets, into a single dataset, what minimum provenance fields and validation checks do you require before accepting a record?
Asked once · last asked Aug 17 · Scenario · Senior
How would you deduplicate near-identical records across sources at scale, such as the same experiment reported in two PDFs, without accidentally merging genuinely distinct runs?
Asked once · last asked Aug 17 · Scenario · Senior
Other ways it was asked
- “At scale, would you deduplicate near-identical records across sources, such as the same experiment reported in two PDFs, without accidentally merging genuinely distinct runs?”
How would you write a senior-level grading rubric to evaluate whether a submitted data-set curation pipeline's deduplication is scientifically valid, including criteria for numerical correctness, lineage integrity, and failure-mode handling?
Asked once · last asked Aug 17 · Scenario · Senior
In Power Query, when combining monthly files from a folder, how do you design the query so schema changes, such as extra or missing columns, don’t break refresh and the output stays consistent?
Asked once · last asked Aug 19 · Tooling · Senior
How do you make a Power Query resilient to schema changes over time without constantly editing the sample file, and how do you handle data type enforcement safely?
Asked once · last asked Aug 19 · Tooling · Senior
During an ETL rerun, target row count increases by 5% while source row count is unchanged. What duplicate check would you run first?
Asked once · last asked Aug 20 · Scenario · Mid-level
How would you handle records that fail business-rule validation during a row-by-row data transfer?
Asked once · last asked Aug 22 · Scenario · Senior
How would you handle missing fields and newly introduced unexpected fields in an ETL pipeline?
Asked once · last asked Aug 23 · Scenario · Mid-level
How would you validate that guardrails in an ETL refresh process prevent bad data snapshots?
Asked once · last asked Aug 29 · Scenario · Mid-level
What primary tests and reconciliation steps would you use to determine where the root data quality issues are occurring across the pipeline?
Asked once · last asked Aug 31 · Scenario · Senior
Suppose A and B come from different systems and arrive with a 24-hour lag mismatch. How would you test whether apparent data-quality violations are real or caused by temporal misalignment?
Asked once · last asked Aug 31 · Scenario · Mid-level
Other ways it was asked
- “Suppose A and B come from different systems and arrive with a 24-hour lag mismatch. How would you test whether the violations are real or caused by a joint temporal alignment issue?”
How would you handle a source column name change to prevent it from breaking a data pipeline?
Asked once · last asked Sep 1 · Scenario · Mid-level
When merging two datasets where one system trims strings and the other preserves trailing spaces affecting joins, how do you detect that and reconcile keys without creating false matches?
Asked once · last asked Sep 7 · Scenario · Senior
In a CSV where some rows have embedded commas and unescaped quotes causing column shift, what checks do you use to identify malformed records and reconstruct the correct final row structure?
Asked once · last asked Sep 7 · Scenario · Senior
How do you implement data validation and cleansing processes in a project?
Asked once · last asked Sep 10 · Scenario · Mid-level
What concerns might arise from putting data quality checks in Airflow tasks?
Asked once · last asked Sep 10 · Scenario · Senior
What does the Unique tool do in Alteryx, and how does it work?
Asked once · last asked Sep 16 · Tooling · Junior
How would you handle data conflicts between Salesforce and an external system in an integration?
Asked once · last asked Sep 18 · Scenario · Senior
The transaction batch is complete, but 2% of the transactions reference customers that are missing from the daily customer feed. Would you block the entire batch, quarantine those transactions, or publish them with incomplete customer information? Explain your decision and what controls you would require.
Asked once · last asked Sep 20 · Scenario · Senior
How would you detect and handle duplicate records in a data pipeline?
Asked once · last asked Sep 20 · Scenario · Mid-level
What design assumptions can allow missing records to reach a report even when a scheduled job reports success?
Asked once · last asked Sep 21 · Scenario · Mid-level
If an on-premises data source changes its data type during a pipeline run, how would you handle it?
Asked once · last asked Sep 21 · Scenario · Senior
How do you handle schema or column changes from a vendor in an Informatica job, and what change management process do you follow?
Asked once · last asked Sep 21 · Scenario · Senior
How can you verify that an API extraction is complete when the API does not provide a reliable total record count?
Asked once · last asked Sep 21 · Scenario · Mid-level
How would you verify that a quarantine rule does not incorrectly exclude valid records or edge cases from downstream processing?
Asked once · last asked Sep 21 · Scenario · Mid-level
What permanent control would you add to prevent a data pipeline from reporting success after retrieving only the first page of an API response?
Asked once · last asked Sep 21 · Scenario · Mid-level
How would you design data quality checks across ingestion, transformation, and publication?
Asked once · last asked Sep 21 · Scenario · Mid-level
Other ways it was asked
- “How have you designed data quality checks across ingestion, transformation, and publication?”
What is an example of a data quality failure that should block a pipeline versus one that should generate only a warning, and how would you decide which response is appropriate?
Asked once · last asked Sep 21 · Scenario · Mid-level
Other ways it was asked
- “Give an example of a data quality failure that should block a pipeline and one that should generate only a warning. How would you decide which response is appropriate?”
If an API does not provide a reliable total record count, how would you verify that all pages were retrieved before republishing a report?
Asked once · last asked Sep 22 · Scenario · Senior
How would you verify that downstream data models use the same completed batch, and what would you block from publication?
Asked once · last asked Sep 22 · Scenario · Senior
How do you design data quality controls across the pipeline, and how do you decide which failures should block publication versus only generate a warning?
Asked once · last asked Sep 22 · Scenario · Senior
How would you ensure that reloading data does not break downstream aggregates and that final tables remain consistent?
Asked once · last asked Sep 22 · Scenario · Senior
How do you integrate data from multiple sources into a CMDB and maintain data quality?
Asked once · last asked Sep 23 · Scenario · Mid-level
Suppose the source does not provide a total record count. How would you decide whether a backfill is complete?
Asked once · last asked Sep 23 · Scenario · Mid-level
How would you explain to an operations manager why a green pipeline does not necessarily mean a dashboard is ready to use, without getting lost in technical details?
Asked once · last asked Sep 23 · Scenario · Mid-level
How would you enforce a schema and convert files before loading them into BigQuery?
Asked once · last asked Sep 24 · Tooling · Mid-level
Why is there a need for daily schema enforcement if the schema is static?
Asked once · last asked Sep 24 · Concept · Mid-level
Other ways it was asked
- “How would you handle occasional dynamic schema changes without running schema enforcement daily?”
Why validate or enforce a schema before loading data into BigQuery instead of handling validation during the transformation stage in an ELT pattern?
Asked once · last asked Sep 24 · Concept · Mid-level
What action do you take when a data quality test fails in your pipeline?
Asked once · last asked Sep 24 · Scenario · Mid-level
How do you handle failed records during validation when loading into BigQuery?
Asked once · last asked Sep 24 · Scenario · Mid-level
What should happen when pre-load tests pass or fail in an ELT pipeline?
Asked once · last asked Sep 24 · Scenario · Mid-level
What do you do when a data ingestion test or validation fails?
Asked once · last asked Sep 24 · Scenario · Mid-level
How do you identify and handle specific malformed records, like addresses with special characters, without discarding the whole batch?
Asked once · last asked Sep 24 · Scenario · Mid-level
Can you explain an example of a data validation rule used in a data pipeline?
Asked once · last asked Sep 24 · Concept · Mid-level
When automating sales reports to reduce manual effort, how do you ensure the accuracy and reliability of the data presented to leadership?
Asked once · last asked Sep 25 · Scenario · Mid-level
An ADF activity returns success, but what would tell you that the extraction stopped too early? How would you change the extraction logic and tests so the same problem is detected before the data reaches Power BI?
Asked once · last asked Sep 28 · Scenario · Senior
If a source legitimately includes summary rows, how would you define and use control totals? Where would you keep those rows, and how would you prevent them from entering the fact table while retaining the evidence needed for reconciliation?
Asked once · last asked Sep 28 · Design · Senior
Suppose a product’s warehouse location or cost differs between ERP and WMS. How would you align the metadata keys for a dimensional model, and what would you do with records you cannot match confidently?
Asked once · last asked Sep 28 · Scenario · Senior
If a financial-crime transaction feed has 100,000 records in both the source and curated dataset, but the total transaction amounts differ by 8%, how would you investigate, and what evidence would you need before allowing downstream users to rely on the dataset?
Asked once · last asked Sep 29 · Scenario · Senior
If a source export contains summary rows mixed in with detail-level data, what permanent control would you implement, where would you classify those rows, what test would block them, and how would you prove the control works without losing valid data?
Asked once · last asked Sep 29 · Scenario · Senior
If a new validation rule quarantines 50 rows, how would you prove they contain no valid customer orders before release? What checks would you run, and what would you do if even one row were ambiguous?
Asked once · last asked Sep 29 · Scenario · Senior
If raw data has 10,000 shipment IDs but the silver layer has 9,000, how would you prove which missing IDs were intentionally excluded and which were lost because of a transformation error? What evidence would you need before releasing the report?
Asked once · last asked Sep 29 · Scenario · Senior
If 800 of 1,000 missing shipment IDs have documented exclusion reasons and the other 200 disappeared after a join to the customer dimension, what would you change in the model, and would you release the report with those 200 shipments missing?
Asked once · last asked Sep 29 · Scenario · Senior
If two source systems have the same number of transactions but their total amounts differ by 8%, what reconciliation would you run to distinguish changed values, offsetting amounts, missing or duplicate records, currency treatment, and timing differences?
Asked once · last asked Sep 29 · Scenario · Senior
Tell me about a time a technically successful data pipeline still produced an untrustworthy business result. What did you initially miss, and what did you change so the same failure would be detected earlier?
Asked once · last asked Sep 29 · Behavioral · Mid-level
If a recovered batch has matching transaction counts between source and curated tables but the total amount differs by 8%, how would you isolate the cause and prove the corrected amount before release?
Asked once · last asked Sep 29 · Scenario · Senior
How do you handle schema changes in SAP Datasphere when a replicated standard CDS view is extended with new fields?
Asked once · last asked Sep 30 · Scenario · Senior
What happens to previously loaded data in a target table when adding a new field to an existing replication flow in SAP Datasphere?
Asked once · last asked Sep 30 · Concept · Senior
How would you ensure that shipment numbers and transfer keys match across systems, especially when discrepancies occur?
Asked once · last asked Oct 7 · Scenario · Senior
Tell me about a time a schema or data contract change silently broke downstream consumers. How did you handle it?
Asked once · last asked Oct 23 · Behavioral · Senior
How would you validate data reconstructed from raw landing data during backdated reprocessing of a previously real-time pipeline?
Asked once · last asked Oct 23 · Scenario · Senior
How would you verify that data is processed in the correct order?
Asked once · last asked Apr 28 · Scenario · Mid-level
Where in a data pipeline would you run jobs that fill gaps in missing data?
Asked once · last asked Jul 20 · Design · Mid-level
How do you identify if the data is the latest or not when moving it between layers?
Asked once · last asked Aug 25 · Scenario · Mid-level
How do you validate source-to-target data?
Asked once · last asked Sep 29 · Concept · Junior
How do you manage data validation and integration?
Asked once · last asked Jan 9 · Scenario · Mid-level

How this list is made

Every question here was asked by a real interviewer in a real job interview that a candidate ran MeetAssist in. After the call, the interviewer's questions are pulled from the transcript. Practice sessions, tests, exams and anything that was not a job interview are dropped.

Questions are rewritten so they make sense on their own: names of people, employers and clients are removed, and speech-to-text mistakes are fixed. Small talk, logistics and questions about one candidate's own résumé are left out. Rephrasings of the same question are merged, and the count is the number of different interviews it came up in.

The interviews come from candidates who use MeetAssist, so they lean towards software, data and AI roles: the shares describe these interviews, not every ETL and Data Pipelines interview. Answers to the most asked questions are written and checked by an editor; a "common mistake" is shown only when reviews of real answers in these interviews show it, with the count.

307 interviews, 169 candidates, Nov 2025 to Oct 2026. Updated every month with new interviews.

All 96 questions

How do you ensure data quality and reconciliation when integrating multiple sources into a reporting pipeline?

How would you handle missing fields in data scraped from a website?

When importing database rows into an application, how would you generate a unique key and validate missing values?

How would you address data integration and data quality risks in a financial dashboard?

How do you design data-quality or integrity tests for data pipelines?

How would you validate that all rows from a source have arrived in a data warehouse?

How should a data pipeline handle a schema mismatch: continue processing with the existing schema, or stop and report an error?

When extracting and transforming data for AI model training, what are your key considerations for data quality, consistency, and security?

When you validate incoming data in Python, what checks do you treat as hard failures versus quarantined bad records, and how do you route or report them?

In Python, how would you implement a reusable data quality framework with rules, severities, and metrics that works across multiple pipelines while still supporting dataset-specific checks?

How would you validate an ETL pipeline end to end without relying on developers?

In production, how do you handle schema evolution, such as added or renamed fields and type changes, across ingestion, transformation, and downstream consumers without breaking the pipeline?

If parent and child records can arrive independently, how would you build a database framework or report to identify child records that have no corresponding parent?

How would you calculate checksums during a data transfer, and what would you do if they did not match because a file was corrupted?

How would you handle integration errors and duplicate data?

How would you monitor data quality in pipelines, and what checks or alerts would you set up to catch errors or anomalies before data reaches end users?

How would you address data quality issues and pipeline failures?

How do you handle schema changes when consuming CDC events into a lake destination?

How would you handle added, removed, or renamed columns when applying CDC events to Delta Lake?

How would you ensure data integrity and engineer robust features when integrating data from sources with varying data quality and schema inconsistencies for a predictive maintenance model?

How do you validate the accuracy of extracted content when passing PDF text to AI, and why might you choose not to use PDF-to-text libraries or IBM Docling?

What tricky edge cases can arise with type conversions or schema checks in a data pipeline, and how would you handle them?

How does an end-to-end sales validation workflow process incoming records, apply validation logic, and produce its outputs?

How do you validate that trading data maps accurately from a trading system to regulatory reporting agents?

When you find anomalies caused by upstream systems rather than your pipeline, how do you decide whether to correct them in your layer or escalate them, and how do you prevent silent fixes from masking source issues?

How do you design data quality gates, thresholds, and pass/fail criteria so a refresh can automatically block or alert when anomalies spike?

When you receive a refreshed extract each day, how would you set up data quality checks that detect schema drift, missing or renamed columns, type changes, and fail the pipeline with clear diagnostics?

How do you ensure data consistency and correctness across the bronze, silver, and gold layers, especially when source schemas change frequently or source files are corrupted?

How would you handle frequent schema evolution in Azure Synapse Analytics without breaking existing queries or pipelines, and what features or techniques would you use to ensure seamless schema evolution?

In an ML data pipeline, what concrete checks would you add to detect training-serving skew early, and where would you enforce them in code?

When curating large volumes of research artifacts, such as PDF tables and analysis spreadsheets, into a single dataset, what minimum provenance fields and validation checks do you require before accepting a record?

How would you deduplicate near-identical records across sources at scale, such as the same experiment reported in two PDFs, without accidentally merging genuinely distinct runs?

How would you write a senior-level grading rubric to evaluate whether a submitted data-set curation pipeline's deduplication is scientifically valid, including criteria for numerical correctness, lineage integrity, and failure-mode handling?

In Power Query, when combining monthly files from a folder, how do you design the query so schema changes, such as extra or missing columns, don’t break refresh and the output stays consistent?

How do you make a Power Query resilient to schema changes over time without constantly editing the sample file, and how do you handle data type enforcement safely?

During an ETL rerun, target row count increases by 5% while source row count is unchanged. What duplicate check would you run first?

How would you handle records that fail business-rule validation during a row-by-row data transfer?

How would you handle missing fields and newly introduced unexpected fields in an ETL pipeline?

How would you validate that guardrails in an ETL refresh process prevent bad data snapshots?

What primary tests and reconciliation steps would you use to determine where the root data quality issues are occurring across the pipeline?

Suppose A and B come from different systems and arrive with a 24-hour lag mismatch. How would you test whether apparent data-quality violations are real or caused by temporal misalignment?

How would you handle a source column name change to prevent it from breaking a data pipeline?

When merging two datasets where one system trims strings and the other preserves trailing spaces affecting joins, how do you detect that and reconcile keys without creating false matches?

In a CSV where some rows have embedded commas and unescaped quotes causing column shift, what checks do you use to identify malformed records and reconstruct the correct final row structure?

How do you implement data validation and cleansing processes in a project?

What concerns might arise from putting data quality checks in Airflow tasks?

What does the Unique tool do in Alteryx, and how does it work?

How would you handle data conflicts between Salesforce and an external system in an integration?

How would you detect and handle duplicate records in a data pipeline?

What design assumptions can allow missing records to reach a report even when a scheduled job reports success?

If an on-premises data source changes its data type during a pipeline run, how would you handle it?

How do you handle schema or column changes from a vendor in an Informatica job, and what change management process do you follow?

How can you verify that an API extraction is complete when the API does not provide a reliable total record count?

How would you verify that a quarantine rule does not incorrectly exclude valid records or edge cases from downstream processing?

What permanent control would you add to prevent a data pipeline from reporting success after retrieving only the first page of an API response?

How would you design data quality checks across ingestion, transformation, and publication?

What is an example of a data quality failure that should block a pipeline versus one that should generate only a warning, and how would you decide which response is appropriate?

If an API does not provide a reliable total record count, how would you verify that all pages were retrieved before republishing a report?

How would you verify that downstream data models use the same completed batch, and what would you block from publication?

How do you design data quality controls across the pipeline, and how do you decide which failures should block publication versus only generate a warning?

How would you ensure that reloading data does not break downstream aggregates and that final tables remain consistent?

How do you integrate data from multiple sources into a CMDB and maintain data quality?

Suppose the source does not provide a total record count. How would you decide whether a backfill is complete?

How would you explain to an operations manager why a green pipeline does not necessarily mean a dashboard is ready to use, without getting lost in technical details?

How would you enforce a schema and convert files before loading them into BigQuery?

Why is there a need for daily schema enforcement if the schema is static?

Why validate or enforce a schema before loading data into BigQuery instead of handling validation during the transformation stage in an ELT pattern?

What action do you take when a data quality test fails in your pipeline?

How do you handle failed records during validation when loading into BigQuery?

What should happen when pre-load tests pass or fail in an ELT pipeline?

What do you do when a data ingestion test or validation fails?

How do you identify and handle specific malformed records, like addresses with special characters, without discarding the whole batch?

Can you explain an example of a data validation rule used in a data pipeline?

When automating sales reports to reduce manual effort, how do you ensure the accuracy and reliability of the data presented to leadership?

An ADF activity returns success, but what would tell you that the extraction stopped too early? How would you change the extraction logic and tests so the same problem is detected before the data reaches Power BI?

If a source legitimately includes summary rows, how would you define and use control totals? Where would you keep those rows, and how would you prevent them from entering the fact table while retaining the evidence needed for reconciliation?

Suppose a product’s warehouse location or cost differs between ERP and WMS. How would you align the metadata keys for a dimensional model, and what would you do with records you cannot match confidently?

If a financial-crime transaction feed has 100,000 records in both the source and curated dataset, but the total transaction amounts differ by 8%, how would you investigate, and what evidence would you need before allowing downstream users to rely on the dataset?

If a source export contains summary rows mixed in with detail-level data, what permanent control would you implement, where would you classify those rows, what test would block them, and how would you prove the control works without losing valid data?