BI Importer

BI tools share common structures. We have defined a format so you can load your dashboard metadata into Catalog. Send a CSV file as described below to our endpoint using the Catalog Uploader.

API Token

Catalog administrators generate API tokens in Settings > API. See Getting Your Catalog API Keys for creation steps and rotation guidance when a token expires or changes.

File Naming

Always prefix file names with a Unix timestamp.

File Expected Content

If you build these files in Excel or Google Sheets, save as CSV (Comma delimited), not CSV UTF-8.

1. dashboards.csv

dashboards.csv example file

	✅ Required	Description	Example
id 🔑	✅	Id of the dashboard (usually from the API)	hlaAw17G
dashboard_type 🔑	✅	Can be DASHBOARD or TILE. A tile is a dashboard element. A Viz Model is a semantic or modeling layer.	DASHBOARD|TILE<br/>|VIZ_MODEL
name	✅	The Dashboard name	MRR Report
folder_path	✅	The full path of the dashboard in your folder structure (do not include the dashboard name in the path)	/Home/Finance/2021
url	✅	The url of the dashboard (to redirect users from Catalog to it)	https://tableau.com/catalog/mrr_report-hlaAw17G
description	❌	Dashboard description	Compute MRR records for 2021.
user_name	❌	The email of the dashboard owner	arnaud@coalesce.io
view_count	❌	Used to compute popularity	255
parent_tables	❌	Tables used in the dashboard	['dbname.schema.table_1', 'dbname.schema.table_2']
parent_dashboards	❌	If the dashboard is built using another dashboard, dashboard id	['dashboard_id_1', 'dashboard_id_2']

To push dashboard lineage to your warehouse data in BigQuery, Snowflake, Redshift, and other warehouses, provide either the Parent Tables field or the dashboard_queries.csv file.

2. dashboard_queries.csv (optional)

dashboard_queries.csv example file

Column Name	✅ Required	Description	Example
dashboard_id	✅	Relate to the dashboard the query belongs to.	hlaAw17G
dashboard_type	✅	Relate to the dashboard the query belongs to.	dashboard|tile
text	✅	The SQL query	SELECT id FROM my_schema.foo
database_name	✅	The technical name of the database the query was executed on *	dw_prod

Catalog needs the database_name column to resolve ambiguities when table names are not prefixed in the query. In the example above, the complete table name is dw_prod.my_schema.foo.

3. dashboard_fields.csv (optional)

dashboard_fields.csv example file

Column Name	✅ Required	Description	Example
dashboard_id	✅	Relate to the dashboard the query belongs to.	hlaAw17G
external_id	✅	The id of the field in the source	hlaAw17G_28
name	✅	Name of the field	arr_certified
role	✅	Role of the field	[MEASURE, DIMENSION]
data_type	❌	Type of data	Number
description	❌	Field description	Measure of ARR
is_primary_key	❌	Default to False	True | False
label	❌	Label of the field	ARR
view_name	❌	Name of the view the field should be grouped in	financial_metrics
view_label	❌	Label of the view the field should be grouped in	Financial Metrics
child_dashboards	❌	Dashboards using the field	['dashboard_id_1', 'dashboard_id_2']
parent_columns	❌	Columns used in the field	['dbname.schema.table_1.columnA', 'dbname.schema.table_2.columnB']

Dashboard fields are only supported for dashboards with dashboard_type = VIZ_MODEL.