Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.datagenie.ai/llms.txt

Use this file to discover all available pages before exploring further.

Use this page as a quick reference. Terms are grouped by where you’ll encounter them most.

​
Data model

A defined collection of metrics and dimensions that DataGenie monitors. Every autonomous insight is scoped to a dataset.
A quantity you track over time — revenue, conversion rate, API latency, contribution margin. Each KPI has a definition, an SQL expression, and a time granularity.
A way to slice a KPI: country, channel, device, customer, product family. Dimensions are how where gets answered.
The time column DataGenie uses for temporal analysis — typically an order date, event date, or ingestion timestamp.
The aggregation level for time-based analysis — hourly, daily, weekly, monthly, quarterly. GO selects the lowest meaningful granularity per KPI.
GO’s output: the complete operational definition of a dataset — KPIs, dimensions, SQL, transforms, granularity. Reusable and exportable.

​
Stories and attribution

A connected anomaly narrative with the structure When · Where · What · What Else · Why. Ranked by business impact.
The underlying KPI that explains a Top Story’s headline change (e.g., “Payment Completion Rate -21%” behind a “Revenue -11%” headline).
The 2–3 dimension values that explain most of a Root KPI’s movement (e.g., “PayPal, Mobile, iOS”).
Dimensions that moved against the primary change — useful for understanding which segments held up.
Related KPIs that moved alongside the primary one — often surfaced via Nirvana across multiple sources.
The ranking score that determines where a story appears in Top Stories. Aligned to the KPIs in your active View Filter.

​
Detection and models

A named configuration of detection parameters — baseline models, seasonality handling, thresholds — applied to a set of KPIs.
Configure multiple AD Groups in parallel and compare their predictions side-by-side in Explorer, Top Stories, and Deep Dive.
Top Stories layout that prioritizes a Hero KPI and groups related KPIs by relevance — powered by deterministic prioritization.
A saved configuration that controls which stories get generated — KPI filters, dimension filters, sub-population filters, and depth filters. Not a reporting filter.

​
Wisdom (agentic analytics)

DataGenie’s agentic analytics assistant. Plans, executes, and explains multi-stage analytical questions in plain English. Not text-to-SQL.
Natural-language rules that encode your business logic — definitions, exclusions, aliases. Wisdom answers within these rules.
A packaged, reusable analytical pipeline defined in plain language (e.g., “Margin Risk Identifier”). Invocable with arguments, output is deterministic.
A documented timeline entry — promotions, outages, seasonality — that Wisdom uses to explain observed patterns with recorded causes, not invented ones.
Pre-configured analytical patterns — forecasting, change analysis, segment comparison, anomaly explanation — each backed by a deterministic service.

​
Platform

DataGenie’s algorithm for virtually joining data across sources at the aggregate level — no ETL, no fact-table joins, no double-counting.
Natural-language dataset onboarding. Describe the business problem; GO returns the Blueprint.
A dashboard promoted to the home screen for always-visible pulse on key insights.
DataGenie’s email-delivery alert channel for Top Stories and scheduled digests.

​
What’s next

Core concepts

The same ideas, in a narrative walkthrough.

Quickstart

The fastest path to your first insight.