Natural language to SQL and data notes
With a known data dictionary, turn business questions into executable SQL with filters, time windows, and metric notes; flag assumptions and expensive scans that need a DBA review.
Case category · Data & analytics
Data & analytics
5 cases Category 5 of 20
This band spans querying, dashboard readouts, spatial analysis, spreadsheet–database sync, and behavioral funnels. Agents can draft SQL and metric commentary under governed data access; execution still requires permissions and audit. Align metric definitions with Experiments & insight cases before comparing numbers across teams.
In the case hub it is Data & analytics (#cat-data), next to Experiments & insight: this band leans on routine analytics and monitoring; the insight band leans on experiment design, features, and data-quality methodology.
Quick links
Natural language to SQL and data notes
Schema, filters, aliases, metric definitions.
Business metrics dashboards and anomaly readouts
Metric definitions, trends, anomaly attribution, actions.
GIS spatial analysis and layer scripts
CRS, buffers, layer overlays, query examples.
Excel / tables synced with databases
Field mapping, incremental sync, unique keys, conflict handling.
Funnel, retention, and user path analysis
Conversion steps, cohorts, paths, funnel diagnosis.
In depth
Business metrics dashboards and anomaly readouts
Explain trend breaks against definitions; separate seasonality, campaigns, and outages; list plausible causes and next drill-down dimensions instead of a single definitive story.
GIS spatial analysis and layer scripts
Handle CRS transforms, buffers, spatial joins, and overlays; include script snippets and topology checks to avoid distance errors from wrong projections.
Excel / tables synced with databases
Design field maps, primary/unique keys, and incremental strategies; document conflict rules (source of truth) and batch transaction boundaries for ops tables and dimensions.
Funnel, retention, and user path analysis
Define funnel steps and windows, quantify step drop-off and cohort retention, and use path analysis to spot unexpected jumps—fueling product and growth experiments.