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.

Traditional BI tools show you what happened. You build charts, configure alerts on a handful of metrics, and hope a human notices when something drifts. DataGenie inverts that model: it autonomously monitors your data at every meaningful slice and surfaces what changed — and why — before you ask.

​
The scale problem

Your business runs on more than the dozen metrics on a dashboard. Every KPI multiplied by every dimension, every dimension value, every sub-population — that’s where the real signal lives. No human, and no reactive BI tool, can manually track it all.
Most analytics tools
  • Track 10–50 metrics you configured
  • Alert on thresholds you guessed at
  • Surface isolated anomalies
  • Require a human to stitch the story together
DataGenie
  • Tracks 100,000s of KPI × dimension combinations automatically
  • Detects anomalies proactively, no thresholds to tune
  • Connects related anomalies into one story
  • Delivers the why, not just the what
From just 6 KPIs across 6 dimensions, DataGenie can auto-create and track well over 100,000 metric combinations — and do it at the time granularity that actually surfaces signal (hourly for SLAs, weekly for marketing trends, monthly for fiscal reporting).

​
Anatomy of an autonomous story

Every anomaly DataGenie surfaces is packaged as a connected story — not an isolated alert. The structure is consistent and scannable:

When

The time period where the change was detected (e.g., “Yesterday”, “Week ending 2026-04-18”).

Where

The dimensional context — region, segment, product family, cohort (e.g., “in US, on iOS”).

What

The primary KPI anomaly — the headline change (e.g., “Revenue dipped 11%”).

What else

Related KPIs that debunk the obvious explanations (e.g., “Visitors and bag-rate were up — so traffic wasn’t the issue”).

Why (Root KPI + Contributors)

The underlying driver — the Root KPI that changed, and the 2–3 dimension values that explain most of the movement (e.g., “Payment Completion Rate fell 21%, driven by PayPal on Mobile/iOS”).
This structure means a director or VP can read a story in 15 seconds and know what to do next — no SQL, no dashboard-hunting, no analyst hand-off.

​
How detection works

DataGenie runs an ensemble of deterministic algorithms — seasonality models, trend models, and low-data models — and selects the best forecast per KPI. A custom seasonality detector (Fourier transforms + autocorrelation validation) handles irregular cycles that off-the-shelf models miss. Everything in this pipeline is deterministic. There are no generative models inventing numbers, no LLM-authored root causes. Learn more in Responsible AI.

​
Where you see Autonomous Insights

Top Stories

Impact-ranked feed of connected stories, updated every processing cycle.

KPI Attribution

Tree-based view of Contributors, Opposers, and Side Effects for any story.

Alerts

Push insights to Slack, Teams, Email, or JIRA — no dashboards to babysit.

​
What’s next

HyperConnected Insights

See how DataGenie connects insights across multiple data sources — with no ETL.

How DataGenie works

The 2-minute mental model of the full platform.