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.

Most enterprise questions cross data sources. Revenue lives in one system, support tickets in another, product usage in a third. Answering “why did margin fall in the Northeast?” usually means building an ETL pipeline, joining fact tables, reconciling granularities, and keeping it all in sync. DataGenie’s Nirvana algorithm skips that entire layer. It virtually joins your data at the aggregate level — so cross-source insights are deterministic, cheap, and maintenance-free.

​
What Nirvana is

Each table or source is onboarded to DataGenie independently — with its own KPIs, dimensions, and time granularity. Nirvana then creates a virtual dataset that combines KPIs across sources, and runs anomaly detection on the combined view. No fact-to-fact joins. No raw-data preprocessing. No ETL job to maintain.
Conventional multi-source analytics
  • Build and maintain ETL pipelines
  • Join fact tables (granularity mismatch, double-counting)
  • Re-run everything when schemas change
  • Expensive compute and storage
  • Brittle — one bad join breaks the report
Nirvana (HyperConnected)
  • Onboard each source independently
  • Virtual join at aggregate level, no raw joins
  • Granularity differences handled automatically
  • No preprocessing compute or storage overhead
  • Deterministic — no double-counting, no drift
Because Nirvana operates on pre-aggregated metrics, cross-source insights are mathematically deterministic. There’s no surface area for hallucinated joins or unexplained variance.

​
When you need it

Different granularities

One table is daily, another is weekly, a third is monthly. A conventional join is impossible — Nirvana handles it natively.

Multi-system workflows

CRM + billing + product usage + support — tied together without a data warehouse project.

Fast cross-functional insight

Stop waiting for a pipeline build-out. Nirvana makes the join the moment data is onboarded.

Avoiding ETL maintenance

Schema changes don’t break Nirvana. Each source is independent — change it, re-onboard it, done.

​
Example: healthcare operations

A hospital group wants to understand occupancy and revenue together:
  • Admissions (Postgres) — daily, per-facility
  • Invoices (Azure SQL) — daily, per-patient
  • Registrations (Postgres) — hourly, per-department
  • Bed occupancy (Azure SQL) — hourly, per-facility
Traditionally, answering “why is CM% dropping despite occupancy being up?” requires a data engineering project to join these tables. With Nirvana, each source is onboarded independently with its own KPIs and dimensions — and the platform surfaces a connected story the moment it detects correlated movement across them. No JOIN. No ETL. No wait.

​
How it fits the product

Datasets

Each source is a Dataset. Nirvana composes them into a virtual combined view.

Top Stories

Stories auto-connect KPIs across Nirvana-joined sources.

Correlation Matrix

Explore cross-source metric relationships auto-discovered by Nirvana.

​
What’s next

Autonomous Insights

The other half of the superpower pair — automatic detection across millions of combinations.

Responsible AI

Why Nirvana’s aggregate-only design is also a safety feature.