01
Live Dashboards
Sub-second refresh on dashboards backed by petabyte fact tables — no nightly extracts, no cube rebuilds, no stale numbers.
Analytics & reportingGENEDATA / 01
A clearer view.A better decision.
Turn shared business definitions into dashboards, analysis, and reporting. Keep the answer connected to the data behind it.
Shared context · Lineage · Governance
Connected
Your sources
Certified metrics
Business questions
Operational data
Connected intelligenceAnalytics workspace
Governance
Business impactDecision briefings
Shared contextLineageGovernance
02
Natural-Language Query
Ask questions in plain English. The semantic layer ensures answers always come from approved metrics and definitions.
03
Pixel-Perfect Reports
Scheduled, branded, regulator-grade PDFs and Excel exports — including footnotes, signatures, and version history.
Semantic Layer
One definition of revenue, however it is asked for.
Dashboards, natural-language questions, scheduled reports, and embedded analytics all resolve through the same semantic layer. That is why the number in a board deck matches the number in a notebook — they are the same definition executed once, not two implementations that happen to agree.
How It Works
Define, certify, publish, consume.
Metrics are defined once by the people accountable for them, then consumed everywhere without being redefined in each tool.
Define the metric
Express revenue, churn, or margin once in the semantic layer, with the joins, filters, and grain stated explicitly.
Certify the definition
An accountable owner signs off. Certified metrics are visually distinguished from ad-hoc ones everywhere they appear.
Publish once
The definition becomes available to dashboards, natural-language queries, exports, and the embedding API simultaneously.
Consume anywhere
Access policy applies at query time, so the same dashboard shows each viewer only the rows they are permitted to see.
Capabilities
What business users get.
Adoption depends on whether the answer arrives quickly and whether people trust it. Both are design constraints here.
Sub-second refresh
Dashboards read from accelerated paths over live tables, so there are no nightly extracts and no stale figures.
Natural-language query
Plain-English questions resolve against certified definitions, so answers cannot silently invent a new metric.
Certified metrics
Trusted definitions are marked as such, with owner and last review date visible at the point of use.
Regulator-grade export
Branded PDF and Excel output with footnotes, signatures, and version history suitable for filing.
White-label embedding
Embed dashboards into your own product on your own domain, with row-level policy applied per end customer.
Point-in-time reporting
Re-run any report as of a past date and reproduce exactly the figures that were published then.
In Practice
Who this is for.
BI is where the platform meets people who do not think of themselves as data users.
Finance
Close with numbers that tie out
Reports resolve through certified definitions and can be reproduced as of any prior date, so restatements are explainable.
Figures that reconcile without a spreadsheet.
Data Analyst
Publish a metric, not another dashboard
Define a metric once in the semantic layer and have it appear consistently everywhere rather than re-implementing it per report.
Fewer near-duplicate dashboards to maintain.
Product
Ship analytics inside your product
Embed governed dashboards into a customer-facing application with per-tenant row-level policy applied automatically.
Customer-facing analytics without a second stack.
What Changes
What changes with a governed semantic layer.
When every tool defines its own metrics, disagreement is guaranteed and reconciliation becomes a standing meeting.
| Dimension | Before Genedata | With Genedata |
|---|---|---|
| Metric definitions | Re-implemented per dashboard and per tool | Defined once, consumed everywhere |
| Freshness | Nightly extracts and cube rebuilds | Sub-second refresh over live tables |
| Access control | A separate permission model in the BI tool | Platform policy applied at query time |
| Trust | Two dashboards, two numbers, one argument | Certified definitions with a named owner |
| Historical reporting | Restatements are difficult to explain | Any report reproducible as of a past date |
IncrementalRefresh
Plan-scopedConcurrent Users
White-labelEmbed Domains
PDF · XLSX · CSVExport Formats
The next step
One semantic layer. Every metric definition aligned.
When marketing, finance, and operations all read from the same governed metric definitions, alignment stops being a meeting topic. Genedata BI is the visible face of the platform — and the layer where the business actually consumes data.
FAQ
Business intelligence, answered.
What analytics leaders and finance teams ask before standardising on a BI layer.
What is Cortex SQL?
Cortex SQL is the analytics product on the Genedata platform: the semantic layer, dashboards, ad-hoc exploration, natural-language query, and regulator-grade reporting. It reads the governed tables GeneFlow produces and honours GeneCatalog policy at query time, so self-service output is reviewable rather than a shadow dataset.
What does the semantic layer actually change?
Metrics are defined once, with joins, filters, and grain stated explicitly, then consumed by dashboards, natural-language queries, scheduled reports, and the embedding API. That is why a figure in a board deck matches the one in a notebook — they are the same definition executed once, not two implementations that happen to agree.
How is natural-language query kept trustworthy?
Plain-English questions resolve against certified metric definitions rather than generating free-form SQL over raw tables. The model chooses among approved metrics, so it cannot silently invent a new definition of revenue.
Can the same dashboard show different data per viewer?
Yes. Platform access policy applies at query time, so row-level and column-level restrictions narrow results per viewer. One dashboard serves every audience without maintaining a filtered copy per team or per customer.
Can we reproduce a report exactly as it was published?
Yes. Any report can be re-run as of a past date and reproduce the figures published then, which is what makes a restatement explainable to finance and to auditors.
Related content
Specification
Semantic Layer Specification
Defining, certifying, and versioning metrics across the platform.
GuideWhite-Label Embedding
Embedding governed dashboards with per-tenant row-level policy.
GuideRegulator-Grade Reporting
Branded exports with footnotes, signatures, and version history.
RelatedGeneCatalog
The governance layer that decides which rows each viewer sees.
Take the next step
Put every team on the same numbers.
See the semantic layer resolve one metric across a dashboard, a natural-language question, and a scheduled report.