Which combination of steps will meet these requirements with the LEAST operational overhead?

Explanations:

Amazon Athena allows users to run SQL queries directly on data stored in Amazon S3, making it suitable for one-time queries. Amazon QuickSight can then visualize this data to create dashboards for KPIs. This combination minimizes operational overhead as it leverages serverless services without needing extensive setup or management.

Amazon Kinesis Data Analytics is designed for processing streaming data in real time, not for one-time queries on batch data. While it can analyze data as it streams in, it doesn’t fulfill the requirement for running one-time queries on consolidated batch data, making this option less appropriate.

Creating custom AWS Lambda functions to move records to an Amazon Redshift cluster would require significant operational overhead in managing and scaling the Lambda functions, as well as in maintaining the Redshift cluster, which is not ideal for low operational overhead.

While using AWS Glue for ETL processes is useful, loading data into multiple Amazon OpenSearch Service clusters is not necessary for the stated requirement of running one-time queries and staging data in S3. This approach adds unnecessary complexity and operational overhead.

Using AWS Lake Formation with blueprints simplifies the ingestion of data into a data lake. AWS Glue can then crawl, extract, and load the data into Amazon S3 in a format like Apache Parquet, which is efficient for analytics. This approach minimizes operational overhead by automating data management processes.