Building a Custom Dashboard

To build your own dashboard:

• Configure the workspace — add tables and define relationships between them
• Add a dashboard to the workspace and configure data visualization by adding the required widgets and filters

Configuring the Data Model

The data store in the system consists of a list of tables within a workspace. Tables contain the information required for analytics.

To begin, add data to the system. This can be done in three ways:

1. Create a table manually or using an automation script. To create a table manually, go to the Data Model tab, click Add table, and select Create. In the window that opens, enter the table and column names, specify the data type, and optionally define partitioning and sorting, then click Add. This method is suitable for small datasets. You can also create a new table using automation scripts. To do this, add a new script in the workspace Control Panel and create a table using the SQL query or Create table blocks. If you need to work with system tables stored on the server, you can load them into ClickHouse using the Get system table script block. Other methods for working with system tables are not supported.
2. Import tables. To do this, go to the Data Model tab, click Add table, and select Import. You can upload files in .csv format. Specify the table name, select the field delimiter, qualifier, and encoding. Import is convenient for test data analysis or system functionality verification.
3. Connect tables from a storage. Tables are connected to the workspace from a ClickHouse database whose server has been added in the Data Storages section. You can connect tables that were automatically loaded into the ClickHouse database via automation scripts or imported and disconnected from the Data Model. To add a table this way, click Add table and select Connect. If available tables exist in ClickHouse, a modal window with their list will open. You cannot reconnect tables already added to the workspace. This method is typically used for production purposes.

After adding tables, configure the Data Model. This will allow you to use fields from different tables in a single visualization component and ensure correct filtering behavior.

To configure the Data Model:

1. Select the table to be used as the fact table and click + Add link below it. Then select the reference table and click the link icon. You can also create a link by dragging the fact table onto the reference table.
2. In the modal window that appears, specify the fact column (from the first table) to be used for the link and the reference column (from the second table) to link to the fact column.
3. Select the link type:
   • Many to one — one value from the first table corresponds to multiple values from the second table
   • Many to many — one value from the first table corresponds to multiple values from the second table, and one value from the second table corresponds to multiple values from the first table
4. Confirm the link creation.

You can define links between multiple tables at once.

For more information about working with tables, see the Configuring the Data Model section.

Creating a Dashboard

Create a dashboard:

1. Click + Add and select Dashboard.
2. In the window that opens, enter a dashboard name.
3. Click Save.

The new dashboard opens immediately in edit mode.

To save changes and make them available for viewing, publish the dashboard using the Publish button.

To switch to edit mode, click Edit in the top panel. When switching to edit mode for a published dashboard, a draft is automatically created.

For more information about working with dashboards, see the Dashboards section.

Working with Views

Views are a tool for configuring dashboard structure, allowing you to add visualization elements, set up their relationships, and organize flexible navigation.

A view consists of rows and columns that form a grid for placing components. Components can include widgets, placeholders, rows, and dividers.

By default, the view is empty. To start working, click the + Create row button. A column is automatically added to the row. When hovering over a column in the row, buttons appear for adding new rows or columns. The selected column is highlighted in the view and displayed in the hierarchy in the left views panel.

You can edit a view in visual mode or in the YAML editor for more precise configuration.

Using views, you can:

• Add pop-up windows for KPI drill-down
• Configure flexible navigation that allows opening one view inside another using a placeholder or set up transitions via actions (e.g., using the Button widget)
• Manage dashboard structure using a grid of rows and columns
• Create dynamic dashboards whose elements change based on context

For more information about working with views, see the Views section.

Working with Widgets

Widgets are data visualization elements that help display processes, charts, and metrics in dashboards.

The system provides several widget categories:

• Process widgets — Funnel chart, Process Explorer, Sphere of processes
• Charts — Histogram, Donut chart, Combined bar chart, Strip chart, Line chart, Bar chart
• Tables — Pivot table, Table
• KPI — KPI List, Percentage KPI, Indicator
• Auxiliary — Button, Parameter, Filter, Text

To add a widget to a dashboard:

1. Add a row and specify the desired widget.
2. Use the right panel to configure data, display, filtering, and actions.
3. Specify dimensions and measures that must come from the same table or from linked tables.
4. In the measure and dimension parameters, select the column to build the dashboard on and choose a formula from the predefined list or create your own by selecting Customize.
5. Configure display parameters for dimensions, measures, and the widget as a whole. You can set conditions for showing the title, font size, color, and other visual parameters.
6. If needed, in the Filter tab, configure Standard or Single choice filter mode, ignore filters, set a state to limit filtering, or apply a predefined filter. Standard filter mode allows selecting one or multiple values in the widget in Browse mode, hiding the rest. With Single choice filter mode, unselected values become semi-transparent.

For more information about widget configuration, see the Widgets section.

Configuring Script Execution from a Dashboard

Within a single workspace, you can execute scripts directly from a dashboard. To do this, configure actions in widgets, for example, in a Table. Using actions, you can add, delete, or update table rows, as well as perform actions in other systems (e.g., create tasks or send messages) without navigating to scripts in the workspace.

To configure script execution from a Table:

1. In the workspace Control Panel, create a script with the Manual start trigger. Configure expected fields (e.g., text, number, or date), specify whether they are required, and add the necessary blocks for working with tables.
2. Configure an action in the dashboard. Select the Table widget, open the actions configuration tab in the right panel, add an action, and select the action type — execute a script, update a variable, or navigate to a URL.
3. After configuring the action, a button appears in the widget. It becomes active automatically in Browse mode or after selecting rows, depending on the settings. Select the required rows and click the button.

When the script runs, a status message appears in the lower right corner.

You can also configure actions that execute when clicking on widget indicators. For example, for dimensions (except Strip chart) or events and transitions in Process Explorer and Sphere of processes, you can set script execution, URL navigation, or variable update. These actions execute when clicking the indicator in Browse mode. Additionally, you can configure the Button widget to perform a specified action.

For more information about executing actions from dashboards, see the Running a Script from a Dashboard, Actions on Click, and Button sections.

Working with a Dashboard

In Browse mode, users can view ready-made reports. This mode supports click-based filtering and table report export.

Click-based filtering allows drilling down into chart and diagram metrics. Clicking a value in a widget activates a filter, and the dashboard displays data related to the selected metric.

To apply a filter, click the value and confirm the selection with a checkmark.

To remove a filter, click the value and press the cross icon or use the cross in the filter panel. To remove all filters, click the eraser icon.

Browse mode supports various filter types, including filtering by rows, numeric values, and dates. You can filter data by specific values, ranges, text presence, as well as by empty or non-empty data.

You can also apply filtering by an arbitrary attribute, even if that attribute is not used in widgets.

Table report export is available in .xlsx and .csv formats.

For more information about working with dashboards in Browse mode, see the Browse mode section.