Make sense of it

Descriptive Aggregation

Turn raw data into answers about what happened.

Descriptive Aggregation slices, groups, and summarises your structured data to produce the numbers that tell you what is happening across your operation — counts, rates, distributions, trends. Reach for it when you need a live view of a population, not a single record. It powers dashboards, scheduled reports, and any summary measure a stakeholder asks you to 'pull'.

Shape

Operational dimensions

No human in loop

Runs without a person in the path.

Scheduled

Fires on a clock.

Medium data gravity

Holds working state that compounds over runs.

Read-only inbound

Consumes external data; does not write back.

Inputs

structured dataset (database table, data warehouse query result, event log)
selected measures and grouping dimensions
filters and time range
optional segment definitions

Outputs

aggregated table or cross-tab with computed summary measures
time-series or trend series per dimension
distribution breakdowns by segment
summary statistics ready for presentation layer

Mechanism

Slices, groups, aggregates and pivots a structured dataset along chosen dimensions to produce summary measures (counts, sums, rates, distributions, trends) that describe what is or what happened.

Why this is a primitive

Cannot be decomposed: the aggregate-and-pivot operation over a defined dataset is a single computation — pick measures, group by dimensions, compute summary, emit table/series. The interactive cockpit vs scheduled-PDF distinction is a delivery wrapper handled at composition layer (dashboards compose aggregation + interactive surface; automated reports compose aggregation + share-it distribution). The aggregation computation itself is identical.

Where it shows up

EdTech platform — weekly active-learner count by cohort, broken down by module completion rate, surfaced to programme managers

SaaS vendor — monthly churn count by plan tier and acquisition channel, used to orient the retention team's weekly standup

Logistics operator — daily on-time delivery rate by route and carrier, auto-posted to the ops dashboard each morning

Healthcare provider — patient wait-time distribution by clinic and day-of-week, reviewed in monthly performance meetings

Related primitives

Composes with

Faceted Filtering & Browse →Multi-Role Live Operational Surface →Long-Form Prose Authoring →Scheduled Outbound Distribution →Predictive Inference →Explanatory Inference →Cross-Source Synthesis →Corpus Retrieval →

Alternative to

Classificatory Inference →