From a question to a clear decision.
Answers business questions with governed queries, visualizations, decision narratives, and recurring briefings.
The context behind the work.
Your value is the decision an analysis helps someone make. Finding data is only the beginning: the audience needs to know what the metric means, whether the comparison is fair, and where uncertainty changes the conclusion.
A business metric moves unexpectedly
A team sees an increase in cost. Check whether usage, mix, pricing, or the reporting definition changed. Compare relevant segments and periods before presenting a cause, and keep the source and assumptions attached to the result.
A recurring briefing takes too much preparation
Agree the audience and metric definitions, reuse a trusted analytical model, and save the analysis pattern. Spend the next review explaining meaningful changes instead of rebuilding the same data preparation.
A practical path from task to outcome.
Explain a change in model cost or business performance using a trusted, reproducible analysis.
- 01
Discover governed context
Define the decision, audience, time period, and metric before searching for a dataset.
- 02
Assess needs and evidence
Find the relevant cataloged data and check ownership, freshness, access, and the meaning of key fields.
- 03
Publish trusted outputs
Explore with SQL or analytics views. Compare segments and verify that sampled inference logs are not treated as full transaction counts.
- 04
Brief decision owners
Share the result with definitions, source references, and the limits of the analysis so the audience can act with context.
A decision brief with the analysis, source references, and stated assumptions.
Less repeated effort. More useful work.
Explore the habits and platform connections that can make this role easier, more consistent, and easier to collaborate with.
Waiting for someone to locate the right dataset
Start with the catalog and an identifiable owner instead of an informal search through teams.
Recreating analysis for recurring questions
Reuse agreed models and saved analysis patterns for the next reporting period.
Explaining where a number came from
Carry the source and metric definition into the published analysis.
Measure your own improvement
Choose a baseline before you begin. Review these signals with your team; results depend on your data, process, and implementation.
- Time from a business question to a reviewed answer
- Repeat questions answered with an existing trusted analysis
Build confidence with a first task.
Explain a change in model cost or business performance using a trusted, reproducible analysis.
Use AI with judgment
Use AI to frame questions or draft an explanation; inspect the underlying data and calculations before sharing conclusions.
Your practice checklist
0 / 4 completeThe right surfaces. The right people.
Continue into the product, deepen your knowledge, or follow the next role in the handoff.
Go deeper
Technical workbookDocumentationTechnical workbooks are maintained in English. Workspace access and available capabilities depend on your deployment and permissions.
Bring your own workflow.
Explore how these practices could fit your team, your data, and your operating requirements.