Objective: trace the question through the pipeline
After identifying storage and database roles, describe the steps that turn source events into a business answer. A dashboard is the final view, not proof that data was correctly collected or interpreted. State the decision the user wants to make and how fresh the answer must be. A daily staffing review and a live operational alert can require different processing schedules. Choose freshness because of the decision deadline, rather than treating the shortest delay as automatically the most valuable architecture.
Four roles in a documented pattern
Google’s Dataflow documentation gives an analytics pattern in which Pub/Sub ingests messages, Dataflow transforms or aggregates them, BigQuery supports analytical queries, and Looker presents business insights. The roles explain the pattern; they do not mandate that every project use all four products. Dataflow supports both batch and streaming processing. Batch can suit a bounded set of records processed on a schedule, while streaming can support continuously arriving events when the business benefits from fresher results. The transformation still needs a precise rule.
Governance protects meaning and access
Write down what a record represents, its unit, the time it refers to, and who owns corrections. If two stores use different definitions of a completed sale, sending both records faster will not reconcile the definitions. The analytical owner must settle the meaning or keep the difference explicit. Also assign access: IAM controls which principal can perform which action on a resource. An analytics platform’s managed infrastructure does not decide whether a finance analyst should see customer identifiers or only an approved aggregate.
Worked example
Fictional case: a bicycle-share operator needs a view of currently available bikes and a separate daily utilization report. First distinguish the decisions: dispatching needs timely events, while yesterday’s planning report can use scheduled processing. Second, define availability and a stable station identifier so the two outputs refer to the same entities. Third, evaluate a message-ingestion and processing path for the timely view, with BigQuery as an analytical destination where appropriate. Fourth, check an event that arrives late or contains an invalid station identifier. The business must decide how that event affects each result; a green pipeline does not decide its meaning.
Change the requirement, not just the technology
Suppose the operator cancels live dispatch and only needs a report before the next morning’s meeting. Streaming is no longer justified merely by the original operational deadline. Compare a simpler scheduled path against the new freshness requirement. Conversely, converting an existing batch to continuous execution does not automatically solve missing source events or inconsistent timestamps. These are separate data-quality questions. Document the accepted handling rule, assign an owner, and inspect representative output records rather than relying only on the job’s completion status.
Transfer check
A museum receives a fixed export at closing time and needs a visitor-count report by noon the next day. Is continuous streaming established as necessary? Check: no; a scheduled batch is a reasonable starting point if it meets the stated deadline. The report still needs an agreed visitor definition and appropriate access. Retrieve the roles of ingestion, transformation, analysis, and presentation with the cards. Next, carry these data-quality obligations into AI problem framing, where unclear inputs can produce misleading model outputs.
Try it, then explain it.
Complete a short set without looking at the example. After each answer, explain which evidence or rule supports your decision and why the alternatives do not fit.
Sources & authorship
Original OpenExam explanations and examples. References support the concepts; they are not the source of copied questions. Independent educator review is pending.
BigQuery overview | Google Cloud DocumentationDataflow overview | Google Cloud DocumentationIAM overview | Identity and Access Management (IAM) | Google Cloud Documentation