Skip to main content

Core concepts

Six things make up almost everything you'll do in Fanar.

Data source​

A connection to a database, warehouse, API, or file. Admins create data sources; everyone else uses them. Permissions on a data source decide who can query it — including what the AI can see when it answers your questions.

See Connect your data.

Query​

A piece of SQL saved against one data source. Queries have a name, an owner, and optionally a schedule and parameters. Everyone on your team can see and fork each other's queries.

A query is also the unit Fanar's AI works with: it searches your saved queries before writing new SQL, and Copilot edits one at a time.

Visualization​

A chart, table, or other rendering of a query's results. One query can have many visualizations — the same data as a line chart, a table, and a counter.

See Visualizations.

Dashboard​

A collection of visualizations and text arranged on a grid. Dashboards can be shared with your team or published publicly, and can be built or edited conversationally with the Dashboard Copilot.

See Dashboards.

Alert​

A condition on a query's results, checked after each run, that notifies a destination when it's met. Useful for "tell me when this number crosses that line".

See Alerts.

Topic​

A governed model in the semantic layer: metrics and dimensions defined once, in YAML, so questions are answered from curated business definitions instead of raw tables.

Topics are for definitions you want enforced. For definitions you want the AI to follow, see Teaching Fanar your business.

How they fit together​

A data source is queried by a query, which is drawn as one or more visualizations, which are arranged on a dashboard. An alert watches a query. A topic sits in front of a data source and defines what its numbers mean.

The AI features work across all of it: ask a question, get a query and a visualization; ask for a dashboard, get one built.