Since the use of the tool is infrequent/intermittent, you can choose to compute options that are serverless. Both Cloud Run and Cloud App Engine Standard are serverless options that can shut down to zero. Since cost-effectiveness is a requirement, this will not cost anything during the periods it is not used.

https://www.crowdstrike.com/cybersecurity-101/zero-trust-security/
Zero Trust is a security framework requiring all users, whether in or outside the organization's network, to be authenticated, authorized, and continuously validated for security configuration and posture before being granted or keeping access to applications and data

One of the key principles of DevOps is to release changes frequently and in small batches. This helps to reduce the risk of disruption caused by bugs. If a bug is introduced in a small change, it is easier to identify and fix the bug without affecting a large number of users.

=> Cloud Pub/Sub- Cloud Pub/Sub is the best to be the end-point for ingesting large amounts of data. It will grow as required, can stream data to downstream systems, and can also work with intermittently available backends.
=> Cloud Dataflow- supports streaming data and therefore is an appropriate option for processing the data that is ingested.
=> BigQuery- BigQuery also supports streaming data and its possible to do real time ana-lytics on it.
=> DataStudio- DataStudio and Looker are for visualization. They don't have any in-built analysis.
=> Cloud Functions- Cloud Functions is a useful serverless endpoint. However, Pub/Sub is better in this case because it can also retain messages for a set period if it was not possi-ble to deliver it first time.
=>Cloud Dataproc- Cloud Dataproc is used for Hadoop/Spark workloads and won't be a good fit here.

Explanation
Option A is Correct- BigQuery Flex Slots for cyclical workloads that require extra capacity, or for workloads that need to process a lot of data in a short time, and so would be less expensive to run using reserved slots for a short time.