Lakehouse & Architecture interview questions

1. A.an architecture that adds data-warehouse features like ACID transactions and governance directly on low-cost data-lake storage
2. B.a layered design organizing data into bronze raw, silver cleaned, and gold curated tables for progressive refinement
3. C.a data-warehouse pattern for tracking changes to dimension attributes over time, such as keeping history with type-2 rows
4. D.a dimensional modeling pattern for tracking attribute history, with Type 1 overwriting and Type 2 adding new versioned rows

Answer + AI explanation with Pro

1. A.Schema-on-read
2. B.Slowly changing dimension
3. C.data contract
4. D.Lakehouse

Answer + AI explanation with Pro

1. A.Lakehouse — the practice of moving curated warehouse data back into operational systems like CRMs to activate analytics
2. B.Lakehouse — a central repository that stores raw structured and unstructured data at scale on low-cost object storage
3. C.Lakehouse — a data-quality pattern that writes to a hidden branch, runs validations, then publishes only if checks pass
4. D.Lakehouse — an architecture that adds data-warehouse features like ACID transactions and governance directly on low-cost data-lake storage

Answer + AI explanation with Pro

1. A.a design that processes everything as a stream, replaying the log for reprocessing instead of maintaining a separate batch layer
2. B.a lakehouse table continuously appended from a streaming source and incrementally processed, often the bronze layer of a pipeline
3. C.a data-warehouse pattern for tracking changes to dimension attributes over time, such as keeping history with type-2 rows
4. D.a layered design organizing data into bronze raw, silver cleaned, and gold curated tables for progressive refinement

Answer + AI explanation with Pro

1. A.data lineage
2. B.Medallion architecture
3. C.Reverse ETL
4. D.Data contract

Answer + AI explanation with Pro

1. A.Medallion architecture — a lakehouse table continuously appended from a streaming source and incrementally processed, often the bronze layer of a pipeline
2. B.Medallion architecture — a self-contained, governed, and documented dataset owned by a domain team and treated as a first-class deliverable in data mesh
3. C.Medallion architecture — a design that processes everything as a stream, replaying the log for reprocessing instead of maintaining a separate batch layer
4. D.Medallion architecture — a layered design organizing data into bronze raw, silver cleaned, and gold curated tables for progressive refinement

Answer + AI explanation with Pro

1. A.an architecture that adds data-warehouse features like ACID transactions and governance directly on low-cost data-lake storage
2. B.a central repository that stores raw structured and unstructured data at scale on low-cost object storage
3. C.the layered lakehouse design refining data through bronze (raw), silver (cleaned), and gold (aggregated) tables
4. D.a decentralized approach where domain teams own their data as products served through a self-serve platform with federated governance

Answer + AI explanation with Pro

1. A.Schema-on-read
2. B.data contract
3. C.change data capture
4. D.Data mesh

Answer + AI explanation with Pro

1. A.Data mesh — a layered design organizing data into bronze raw, silver cleaned, and gold curated tables for progressive refinement
2. B.Data mesh — the pattern of capturing inserts, updates, and deletes from a source database's log and streaming them into a lakehouse for near-real-time replication
3. C.Data mesh — a precomputed, stored query result that is incrementally refreshed so downstream reads avoid recomputing expensive aggregations
4. D.Data mesh — a decentralized approach where domain teams own their data as products served through a self-serve platform with federated governance

Answer + AI explanation with Pro

1. A.the layered lakehouse design refining data through bronze (raw), silver (cleaned), and gold (aggregated) tables
2. B.a data-quality pattern that writes to a hidden branch, runs validations, then publishes only if checks pass
3. C.a design that runs parallel batch and speed layers and merges their outputs at query time to balance accuracy with low latency
4. D.a central repository that stores raw structured and unstructured data at scale on low-cost object storage

Answer + AI explanation with Pro

1. A.Write-audit-publish
2. B.CDC
3. C.Lambda architecture
4. D.data product

Answer + AI explanation with Pro

1. A.Lambda architecture — a design that runs parallel batch and speed layers and merges their outputs at query time to balance accuracy with low latency
2. B.Lambda architecture — the layered lakehouse design refining data through bronze (raw), silver (cleaned), and gold (aggregated) tables
3. C.Lambda architecture — a layered design organizing data into bronze raw, silver cleaned, and gold curated tables for progressive refinement
4. D.Lambda architecture — an architecture that adds data-warehouse features like ACID transactions and governance directly on low-cost data-lake storage

Answer + AI explanation with Pro

1. A.the pattern of capturing inserts, updates, and deletes from a source database's log and streaming them into a lakehouse for near-real-time replication
2. B.the tracked path of data from sources through transformations to outputs, used for impact analysis, debugging, and governance
3. C.a precomputed, stored query result that is incrementally refreshed so downstream reads avoid recomputing expensive aggregations
4. D.a design that processes everything as a stream, replaying the log for reprocessing instead of maintaining a separate batch layer

Answer + AI explanation with Pro

1. A.Write-audit-publish
2. B.Kappa architecture
3. C.Data mesh
4. D.Data lake

Answer + AI explanation with Pro

1. A.Kappa architecture — a design that processes everything as a stream, replaying the log for reprocessing instead of maintaining a separate batch layer
2. B.Kappa architecture — an architecture that adds data-warehouse features like ACID transactions and governance directly on low-cost data-lake storage
3. C.Kappa architecture — a design that runs parallel batch and speed layers and merges their outputs at query time to balance accuracy with low latency
4. D.Kappa architecture — a central repository that stores raw structured and unstructured data at scale on low-cost object storage

Answer + AI explanation with Pro

1. A.change data capture, the technique of streaming row-level inserts, updates, and deletes from a source database into a pipeline
2. B.a dimensional modeling pattern for tracking attribute history, with Type 1 overwriting and Type 2 adding new versioned rows
3. C.the approach of storing raw data as-is and applying structure only when it is queried, common in data lakes
4. D.an architecture that adds data-warehouse features like ACID transactions and governance directly on low-cost data-lake storage

Answer + AI explanation with Pro

1. A.Data mesh
2. B.Schema-on-read
3. C.CDC
4. D.Lakehouse

Answer + AI explanation with Pro

1. A.Schema-on-read — a layered design organizing data into bronze raw, silver cleaned, and gold curated tables for progressive refinement
2. B.Schema-on-read — the approach of storing raw data as-is and applying structure only when it is queried, common in data lakes
3. C.Schema-on-read — a self-contained, governed, and documented dataset owned by a domain team and treated as a first-class deliverable in data mesh
4. D.Schema-on-read — the layered lakehouse design refining data through bronze (raw), silver (cleaned), and gold (aggregated) tables

Answer + AI explanation with Pro

1. A.a central repository that stores raw structured and unstructured data at scale on low-cost object storage
2. B.a data-warehouse pattern for tracking changes to dimension attributes over time, such as keeping history with type-2 rows
3. C.a lakehouse table continuously appended from a streaming source and incrementally processed, often the bronze layer of a pipeline
4. D.a data-quality pattern that writes to a hidden branch, runs validations, then publishes only if checks pass

Answer + AI explanation with Pro

1. A.change data capture
2. B.Schema-on-read
3. C.Slowly changing dimension
4. D.data product

Answer + AI explanation with Pro

1. A.Slowly changing dimension — a decentralized approach where domain teams own their data as products served through a self-serve platform with federated governance
2. B.Slowly changing dimension — a design that runs parallel batch and speed layers and merges their outputs at query time to balance accuracy with low latency
3. C.Slowly changing dimension — a data-warehouse pattern for tracking changes to dimension attributes over time, such as keeping history with type-2 rows
4. D.Slowly changing dimension — a precomputed, stored query result that is incrementally refreshed so downstream reads avoid recomputing expensive aggregations

Answer + AI explanation with Pro

1. A.the tracked path of data from sources through transformations to outputs, used for impact analysis, debugging, and governance
2. B.a formal agreement on schema, semantics, and quality between data producers and consumers to keep pipelines stable
3. C.a lakehouse table continuously appended from a streaming source and incrementally processed, often the bronze layer of a pipeline
4. D.a layered design organizing data into bronze raw, silver cleaned, and gold curated tables for progressive refinement

Answer + AI explanation with Pro

1. A.Data contract
2. B.medallion architecture
3. C.Data mesh
4. D.streaming table

Answer + AI explanation with Pro

1. A.Data contract — a formal agreement on schema, semantics, and quality between data producers and consumers to keep pipelines stable
2. B.Data contract — an agreed, versioned specification of a dataset's schema, semantics, and quality guarantees between producers and consumers
3. C.Data contract — a dimensional modeling pattern for tracking attribute history, with Type 1 overwriting and Type 2 adding new versioned rows
4. D.Data contract — the approach of storing raw data as-is and applying structure only when it is queried, common in data lakes

Answer + AI explanation with Pro

1. A.a self-contained, governed, and documented dataset owned by a domain team and treated as a first-class deliverable in data mesh
2. B.the practice of moving curated warehouse data back into operational systems like CRMs to activate analytics
3. C.the approach of storing raw data as-is and applying structure only when it is queried, common in data lakes
4. D.a lakehouse table continuously appended from a streaming source and incrementally processed, often the bronze layer of a pipeline

Answer + AI explanation with Pro

1. A.CDC
2. B.materialized view
3. C.Reverse ETL
4. D.data lineage

Answer + AI explanation with Pro

1. A.Reverse ETL — a design that processes everything as a stream, replaying the log for reprocessing instead of maintaining a separate batch layer
2. B.Reverse ETL — change data capture, the technique of streaming row-level inserts, updates, and deletes from a source database into a pipeline
3. C.Reverse ETL — the practice of moving curated warehouse data back into operational systems like CRMs to activate analytics
4. D.Reverse ETL — the tracked path of data from sources through transformations to outputs, used for impact analysis, debugging, and governance

Answer + AI explanation with Pro

1. A.the practice of moving curated warehouse data back into operational systems like CRMs to activate analytics
2. B.a formal agreement on schema, semantics, and quality between data producers and consumers to keep pipelines stable
3. C.a data-quality pattern that writes to a hidden branch, runs validations, then publishes only if checks pass
4. D.the layered lakehouse design refining data through bronze (raw), silver (cleaned), and gold (aggregated) tables

Answer + AI explanation with Pro

1. A.Write-audit-publish
2. B.Schema-on-read
3. C.Kappa architecture
4. D.data contract

Answer + AI explanation with Pro

1. A.Write-audit-publish — a data-quality pattern that writes to a hidden branch, runs validations, then publishes only if checks pass
2. B.Write-audit-publish — a data-warehouse pattern for tracking changes to dimension attributes over time, such as keeping history with type-2 rows
3. C.Write-audit-publish — a central repository that stores raw structured and unstructured data at scale on low-cost object storage
4. D.Write-audit-publish — a design that processes everything as a stream, replaying the log for reprocessing instead of maintaining a separate batch layer

Answer + AI explanation with Pro