Setting up Datasources
Type of Data Sources
A data source can be set up based on a remote database connection, by uploading a file or by connecting to an Api. Data sources can be set up on the Datasource page. Please note that when scanning your database Turbular can trigger multiple read requests on your database.
Supported Databases
Currently, SSH protected databases as well as none SSH protected databases are supported. The following database types are supported:
- MySql
- PostgreSQL
- Oracle
- Microsoft SQL Server
- SQLite
- BigQuery
Saved Database Schemas
Please note that on your first request against your database Turbular scans your database and saves the scanned database schema. If you change
the structure of your database later (for example adding a new table), you will need to resynchronize your database. This can be down via the
Datasource page, simply click on the menu button for the database data source you want to resynchronize. Then in the menu which has opened click
on "Resynchronize".
Supported File Formats
The following file formats are currently supported
- Excel
- CSV
When uploading files make sure that the data types of the file are correct (f.e salary numbers as numbers and not strings) and that the file is not password protected.
Suported Apis
The following Apis are currently supported:
- AWS
- Stripe
- Mailchimp
- Shopify
- Salesforce
- Trello
- Zendesk
- Jira
- Hubspot
- ClickUp
If you feel like we really need to support a certain Api feel free to submit a request on our Contact page. If you have a custom Api you need for your business feel free to contact our sales team about our enterprise offerings.
Ensure Proper Naming Conventions
Turbular's AI models try to infer the meaning of your data sources columns (and tables) based on their names. Therefore, it is important to ensure that you name columns and tables in a manner that reflects their actual meaning.
Good Examples:
- Sales Value
- Order Date
- Address
Bad Examples:
- Revenue1
- Revenue2
- XY_ZZ234r
Uploading Excel Or CSV Files
When uploading Excel or CSV files make sure that the columns in your file have the right data type. A common issue we have observed is that for example prices are mistakenly saved as text when they should be saved as numbers. Turbular does take your data as is therefore if you encounter strange looking charts or model replies wrong data types in your file may be the reason.
Supported Languages
So far Turbular only supports english as a language for database schemas, while different languages may work it will result in degraded performance. Therefore, for both file data sources make sure all column names are in english. For databases data sources all columns and table need to be in english. For example, you have uploaded a Excel File about sales data in german, naming the columns of the file "Order Date","Order Value","City" is fine, while using german columns like "Bestelldatum", "Verkaufswert", "Stadt" would lead to degraded results.