A. Correct.
A database view encapsulates the join logic, and AWS Glue can run a scheduled ETL job that reads through JDBC and writes directly to S3 in Parquet format.

B. Incorrect.
Exporting CSV and then converting with Lambda adds extra moving parts and is less suitable for large daily analytics exports.

C. Incorrect.
Crawlers discover schema, but running a crawler every day is not required to extract a known view. Manual starts also do not meet the daily automation goal cleanly.

D. Incorrect.
Lambda has runtime and packaging limits that make it a poor fit for large data exports compared with a managed Glue ETL job.

A. Incorrect.
Amazon Redshift does not provide a native WebSocket query interface for this use case.

B. Correct.
The Redshift Data API lets applications run SQL through an HTTPS API without persistent connections or database drivers.

C. Incorrect.
JDBC works, but the application must manage drivers, networking, pooling, retries, and credentials.

D. Incorrect.
S3 Select queries individual objects and is not a replacement for real-time SQL access to Redshift warehouse data.

A. Incorrect.
Staging files and scheduling COPY commands increases latency and operational work.

B. Incorrect.
BI tools generally query Redshift objects, not raw Kinesis stream shards directly.

C. Correct.
Redshift streaming ingestion can read from Kinesis through an external schema and materialized view, giving SQL access for near real-time analytics.

D. Incorrect.
Firehose to S3 plus manual or scheduled loads adds extra steps and is not the lowest-overhead near real-time pattern.

A. Incorrect.
MSK is a managed Kafka service, but it requires producer and consumer pipeline design and is not the simplest SaaS integration service.

B. Correct.
Amazon AppFlow is designed to move data between SaaS applications and AWS services such as S3 and Redshift with managed connectors.

C. Incorrect.
The Glue Data Catalog stores metadata. It does not extract data from SaaS applications by itself.

D. Incorrect.
Kinesis is useful for streaming ingestion but does not provide turnkey SaaS application connectors like AppFlow.

A. Incorrect.
Firehose can deliver to Redshift, but the direct streaming ingestion feature is the lower-overhead fit for Kinesis streams and Redshift analytics.

B. Correct.
Redshift streaming ingestion supports Kinesis Data Streams as a source and makes streaming data queryable with low latency.

C. Incorrect.
S3 plus periodic COPY is a batch pattern and adds latency.

D. Incorrect.
Aurora zero-ETL integrates Aurora databases with Redshift; it is not the ingestion path for Kinesis stream data.

A. Incorrect.
Splitting files and managing separate buckets creates more pipeline and permissions work than governed table permissions.

B. Incorrect.
RDS adds database administration overhead and is not the natural analytics layer for S3 data lake files.

C. Incorrect.
Macie can discover sensitive data, but it does not automatically enforce query-level access permissions.

D. Correct.
Lake Formation provides centralized permissions for data lake tables, including column-level controls that can restrict PII while exposing non-PII fields.

A. Correct.
AWS Glue supports data lake table formats by using the --datalake-formats job parameter, such as iceberg.

B. Incorrect.
Auto scaling controls worker capacity behavior, not table format integration.

C. Incorrect.
Job bookmarks track processed data for incremental jobs; they do not enable Iceberg support.

D. Incorrect.
TempDir configures temporary storage, not the data lake table format.

A. Incorrect.
Polling adds delay and unnecessary invocations.

B. Correct.
Glue event-based triggers can respond to EventBridge events, and S3 object-level events can be captured through CloudTrail data events.

C. Incorrect.
Frequent schedules waste resources and still may not start immediately when the file arrives.

D. Incorrect.
DMS does not target AWS Glue workflows in this way.

A. Incorrect.
Custom Lambda code increases development and maintenance effort for rule management.

B. Correct.
DataBrew supports visual profiling and data quality rulesets for S3 datasets, and jobs can be automated with EventBridge.

C. Incorrect.
An always-on EMR cluster is more expensive and operationally heavy for this requirement.

D. Incorrect.
Step Functions plus custom Lambda validation works, but it is not the least-effort no-code approach.

A. Correct.
The FLEX execution class uses spare capacity and can reduce cost for non-urgent Glue Spark jobs.

B. Incorrect.
STANDARD is appropriate for time-sensitive workloads but is not the cost-optimized choice here.

C. Incorrect.
AWS Glue jobs do not expose a normal EC2 Spot instance type setting like self-managed clusters.

D. Incorrect.
GlueVersion affects runtime features and compatibility, but selecting a newer version alone is not the cost control.