Administrator Documentation

Author workflows in the New Data Preparation Experience for Analytics Builder

The New Data Preparation Experience is a visual, step-by-step way to build datasets in Analytics Builder. Instead of a linear list of tables and fields, you have a workflow canvas that shows:

Each transformation as a step
How data flows from one step to the next
Where you have branches (for more complex workflows)
A real-time data preview that updates dynamically

The New Data Preparation Experience also has the following benefits:

More transformation options, including Pivot, Unpivot, Aggregate, and Append.
Support for advanced functionalities, such as branching workflows and composite datasets (using one dataset as the input to another).

This is the bullhorn icon used for warnings.

Important: New datasets that were created using the New Data Preparation Experience open in the new visual workflow by default. Existing datasets that were created in the legacy experience remain in the legacy data preparation editor and cannot be converted.

All downstream features including visuals, dashboards, and sharing function the same in both new and legacy data preparation experiences.

Create a new dataset

You always see the New Data Preparation Experience when you create a new dataset from a data source.

Watch this video to learn how to create a new dataset from existing data sources, using the fast path, or by opening an existing dataset in Analytics Builder. Review the written steps below.

To create a new dataset:

Navigate to Insights Portal.
Under the Analytics Builder tab, click Analytics Builder. Analytics Builder opens in the Analysis tab by default. From this view you can also see the Favorites, Recent, My Folder, Shared Folders, Dashboards, and Datasets tabs.
Click the Datasets tab to show the list of available assets.
Click New dataset to begin creating a new dataset.
Under FROM EXISTING DATA SOURCES, choose your source. For example, select Brightspace, then click Create dataset.
Select the BDS you want to use and click Edit/Preview data.

This opens a New Data Preparation Experience.

View of the data preparation canvas containing one object, labeled 'accommodationsprofilelog-10_7_5'. That label appears at the top of the page as well.

Create a new dataset using custom SQL

Watch this video to learn how to create a new dataset using custom SQL in Analytics Builder. Review the written steps below.

To create a new dataset using custom SQL

Navigate to Insights Portal.
Under the Analytics Builder tab, click Analytics Builder. Analytics Builder opens in the Analysis tab by default. From this view you can also see the Favorites, Recent, My Folder, Shared Folders, Dashboards, and Datasets tabs.
Click the Datasets tab to show the list of available datasets.
Click New dataset to begin creating a new dataset.
Under FROM EXISTING DATA SOURCES, choose your source. For example, select Brightspace, then click Create dataset.
If you want to build using an existing BDS, you can optionally select the BDS you want to use under Tables.
Do one of the following:
1. Click Use custom SQL to begin writing your SQL query immediately.
2. Click Edit/Preview data to open the SQL editor in the new experience first.

Both paths open the new Data Preparation experience, where you can continue building your datasets using SQL.

The Edit SQL query for Analytics Builder Data Prep, with fields for Custom SQL Name and Custom SQL, and an Output Preview below.

Open an existing dataset

When working with datasets you have already built or that come from BDS, you can quickly open them from the Datasets tab.

To open an existing dataset

Navigate to Insights Portal.
Under the Analytics Builder tab, click Analytics Builder.
Click the Datasets tab.
Search for the dataset name.
Click Use in analysis and select USE IN DATASET.

This opens the New Data Preparation Experience.

Use Fast Path

Authors can begin building visuals immediately after selecting a dataset.

To use Fast Path

Navigate to Insights Portal.
Under the Analytics Builder tab, click Analytics Builder.
Click the Datasets tab to show the list of available datasets.
Click New dataset to begin creating a new dataset.
Under FROM EXISTING DATA SOURCES, choose your source. For example, select Brightspace, then click Create dataset.
Select the BDS you want to use and click Select.
Click Visualize.

This opens the dataset in an Analysis, allowing you to begin building visuals.

To open your dataset in edit mode

From the Analysis, click the dataset name in the top left.
Select Manage datasets.
Click the three-dot menu beside your dataset and select Edit.

This opens the New Data Preparation Experience.

View of the data preparation canvas containing one object, labeled activityexemptionslog_10_7_5. That label appears at the top of the page as well.

Overview of the New Data Preparation interface

Watch this video to learn how to navigate the New Data Preparation Experience in Analytics Builder. Review the written steps below.

When you open a dataset in the new experience, the screen contains four main areas:

Workflow Canvas (center area)
Left Pane (Steps panel)
Configure & Preview Panels (bottom area)
Toolbar (top right)

Workflow Canvas

Analytics Builder Workflow Canvas showing connected steps for filters and joins in a dataset workflow.

The Workflow Canvas shows your dataset as a series of connected steps. Each step represents one action taken on the data. For example:

Input (your starting data source)
Filter
Select Columns
Rename Columns
Join
Aggregate
Publish

This allows you to view the complete dataset workflow from start to finish. You may start with a single data source and apply one or two transformations.

Analytics Builder Workflow Canvas showing dataset transformations with filters selects renames and joins.

The workflow can also support more advanced dataset building. When your logic requires separate transformation paths, the canvas displays branches, allowing each path to evolve independently before being joined back together. The following example shows how the canvas handles more advanced dataset building:

Several Organizational Unit datasets are used as inputs.
Each branch applies its own transformations (Filter, Select, Rename).
The branches are later joined together.

Left Pane

The left pane shows every transformation you can apply, grouped into categories.

Analytics Builder left pane showing Steps menu with transformation options for dataset preparation.

Authors can add a transformation by clicking +Add in the left pane or by clicking + directly on the Workflow Canvas.

Analytics Builder left pane highlighting Add calculated columns option with Add action selected.

Data Preview & Configuration Panels

When you select any step in the workflow, the bottom area of the screen displays two panels: Configure and Preview.

Analytics Builder Configure tab showing Change data type settings for a selected dataset column.

Configure is where you define what the step does and ensure you set the transformation correctly. The Configure tab lets you:

Select which columns to modify
See the original data type
Choose a new data type
Confirm what will change before applying it.

Analytics Builder Preview tab showing transformed dataset rows after a change data type step.

Preview is where you verify the results of the step.

After configuring the transformation, the Preview panel updates and shows:

SessionId is now displayed as Decimal
Visual confirmation that the transformation worked as intended.

The Configure and Preview panels help authors validate the change immediately before moving on. This two-step interaction reduces mistakes and increases confidence in the final dataset.

Toolbar

The toolbar gives you quick access to the most important dataset controls. It always appears at the top of the screen while you build your dataset. The name of the dataset always displays in the top left.

Analytics Builder toolbar with the dataset name highlighted on the left. On the right is a dropdown with SPICE selected, a dropdown with Save & Publish selected, and a Cancel button.

Edit Dataset Name

Click the dataset name to rename it. Renaming updates it everywhere the dataset is used.

Query Mode Detector

The toolbar shows whether your dataset is using SPICE or Direct query. This choice affects which transformations are available.

Analytics Builder toolbar with Query Mode Detector open, showing the two selectable options: SPICE and Direct query.

Query Mode: SPICE vs DirectQuery

Use SPICE when you want the full capabilities of the New Data Preparation Experience.

SPICE allows transformations that reshape or summarize your data, including:

Pivot
Unpivot
Aggregate
Append
Complex or multi-step workflows (including branches)

SPICE loads your data into a fast, in-memory engine. This lets Analytics Builder perform heavy operations that cannot run efficiently against live database queries.

Important: If you add any of the above transformations, your dataset becomes SPICE-only and you won't be able to select the DirectQuery option as shown in the example below.

Query Mode Detector open with SPICE selected and Direct query unselectable. The warning message below says: Direct query is unavailable because your dataset includes a data source or transformation that is supported only in SPICE.

DirectQuery supports basic, lightweight operations including:

Filters
Select, Rename, Change Data Type
Basic Calculated Fields
Simple Joins (with some limits)

DirectQuery does not support Pivot, Unpivot, Aggregate, Append, Complex or multi-branch workflows.

Save & Publish

Analytics Builder toolbar with Save & Publish dropdown open, showing the two selectable options: Save & Publish, and Publish & Visualize.

The toolbar includes two publishing options, depending on what you want to do next.

Click Save & Publish to make your dataset available for analyses and dashboards.
Click Publish & Visualize to publish the dataset and open it as a new analysis so you can start building visuals right away.

Important: Once you publish a dataset created in the New Data Preparation Experience, it cannot be edited using the legacy data preparation experience. All future edits will always open in the new experience.

Cancel

Click Cancel to end your current editing session and return to the Datasets tab.

Transform datasets using the New Data Preparation Experience

The new Data Preparation Experience lets you build datasets using a visual, step-by-step workflow. Each transformation appears as a tile on the canvas, and the Preview tab in the bottom panel shows the results immediately.

Below is an overview of how to add data, apply transformations, and use more Advanced Capabilities.

Add data

Every dataset begins with an input step. To add a data source:

From the Input tile, click Add Data.
Select one of the following: Dataset (for custom datasets), Data source, (for Brightspace Data Sets), or Upload a file.

Add data popup with the Select dropdown open, showing the options: Dataset, Data source, and Upload a file.

Common transformations

Watch this video to learn how to use common transformations such as Add calculated columns, Filter, Select columns, and Rename columns steps in Analytics Builder. Review the written steps below.

Filter

Use Filter to keep only the rows that match specific conditions. The Filter narrows your dataset based on rules you define (for example: Keep only Sections, only Active users, only current semesters).

From the left pane, click Filter. You can also click the + icon on the canvas and choose Filter.
In the Configure panel below, choose a column, pick a condition and enter a value.
Verify the filtered results in the Preview panel.

Analytics Builder workspace with Filter Sections block and filter condition set to type equals Section.

Filter is supported in both SPICE and DirectQuery.

Select columns

Use Select columns to choose which fields you want to keep in your dataset. This step removes any irrelevant columns that you don’t need.

From the left pane, click Select columns. You can also click the + icon on the canvas and choose Select columns.
In the Configure panel below, select the columns you want to include.
Reorder the columns if needed.
Use the Preview panel to confirm that only the selected fields are visible.

Analytics Builder view with Select Sections block and three columns selected in the configuration pane.

Select columns is supported in both SPICE and DirectQuery.

Rename columns

Use Rename columns when you want to give field names clearer, more meaningful titles before using them in visuals or joining them with other data sources.

From the left pane, click Rename columns. You can also click the + icon on the canvas and choose Rename columns.
In the Configure panel below, select the columns you want to rename.
Enter the new names under New Name.
Use the Preview panel to confirm that the field names have changed.

Analytics Builder with Rename Sections block showing orgunitid, name, and code renamed to new column names.

Rename columns is supported in both SPICE and DirectQuery.

Change data type

Use Change data type when you need a column to behave as a different data type. For example, you may need to convert a numeric field stored as String into an Integer or Decimal.

From the left pane, click Change data type. You can also click the + icon on the canvas and choose Change data type.
In the Configure panel below, select the columns you want to modify.
Use the NEW DATA TYPE dropdown to select the new type.
Use the Preview panel to confirm the dataset contains the new data types.

Analytics Builder showing Change data type step with StatusType column selected and New data type dropdown open.

Add calculated columns

Use Add calculated columns when you want to create new fields based on formulas. For example, you may need to calculate a duration using a start time and an end time.

Analytics Builder with Add calculated columns step selected and no calculated columns added yet.

From the left pane, click Calculated columns. You can also click the + icon on the canvas and choose Calculated columns.
In the Configure panel below, click + Calculated column.
Use the formula editor to name the new column and build an expression using fields, parameters, and supported functions.
Use the Preview panel to confirm that the new column appears and behaves as expected.

Calculated columns is supported in both SPICE and DirectQuery.

Join

Watch this video to learn how to use the Join transformation in Analytics Builder. Review the written steps below.

Use Join when you want to combine two tables based on matching values to bring related information together into a single dataset.

In the example below, the author joins the Sections branch with the Offer & Parent branch using SectionId = orgunitid.

Analytics Builder with Offer and Section Join step using SectionId and orgunitid as join keys for a right join.

From the left pane, click Join. You can also click the + icon on the canvas and choose Join.
In the Configure panel below, select a column from the Left table.
Select a matching column from the Right table and add more clauses if needed.
Select a Join type:
- Inner
- Left
- Right
- Full
Use the Preview panel to verify that the Join behaves as expected.

DirectQuery supports simple joins only. SPICE supports all join types and multi-step joins.

Append

Watch this video to learn how to use the Append transformation to combine datasets in Analytics Builder. Review the written steps below.

Use Append to combine two datasets at a time by stacking their rows into a single output. If you need to combine more than two datasets, you can Append additional steps to include each dataset sequentially.

Analytics Builder with Append step showing unmatched columns between UserLogins and OrgStructureBeta1 datasets.

From the left pane, click Append. You can also click the + icon on the canvas and choose Append.
In the Configure panel below, select Table 1 (the first dataset) and Table 2 (the dataset you want to append).
Review the Append Output preview to confirm:
- Matching columns
- Columns that exist only in one dataset
- Column alignment
Adjust your source datasets by renaming or selecting columns as needed to ensure the structures match.
Use the Preview panel to verify the rows have been combined correctly.

Append is a SPICE-only transformation. If you add an Append step, the dataset automatically becomes SPICE-only and you will no longer be able to switch to DirectQuery.

Aggregate

Watch this video to learn how to use the Aggregate, Pivot, and Join transformations to summarize data in the New Data Preparation Experience. Review the written steps below.

Use Aggregate when you want to summarize your data. For example, counting records, finding min/max values, or grouping values to produce a higher-level result. In the example below, the author groups by UserName and aggregates StatusType using the Max function to produce a single status value per user.

Analytics Builder with Aggregate step grouping by UserName and using Max on StatusType to create Status column.

From the left pane, click Aggregate. You can also click the + icon on the canvas and choose Aggregate.
In the Configure panel below:
1. Add one or more Group by columns (for example, UserName).
2. Add one or more Aggregated columns (for example, StatusType).
3. Choose an aggregation function such as Count, Count Distinct, Min, Max, or ListAgg.
4. Provide an output column name.
Use the Preview panel to verify the summarized result.

Aggregate is a SPICE-only transformation. If you add an Aggregate step, the dataset automatically becomes SPICE-only and you will no longer be able to switch to DirectQuery.

Pivot and Unpivot

Watch this video to learn how to use the Unpivot transformation in Analytics Builder. For the Pivot transformation, refer to the video located in the Common Transformations section. Review the written steps below.

Use Pivot and Unpivot to reshape data between row-based and column-based formats. These transformations are often used when preparing datasets for analysis or when working with attribute-style data.

Pivot converts row values into columns.
Unpivot converts columns into rows.

From the left pane, click Pivot. You can also click the + icon on the canvas and choose Pivot.
In the Configure panel below:
1. Under Pivot column, select the column whose values will become new column headers.
2. Under Value column, select the column that contains the values to populate the pivoted columns.
3. Under Aggregation function, choose how multiple values are handled.
4. Under Group by columns, select one or more columns to define each output row.
Use the Preview panel to confirm that rows have been converted into columns as expected.

Pivot and Unpivot are supported in SPICE only.

Advanced capabilities

Divergence

Watch this video to learn how to use Divergence in Analytics Builders. Review the written steps below.

The new Data Preparation Experience allows you to split your workflow into parallel branches, each applying its own transformations before merging back together. This is helpful when different parts of your dataset require separate logic.

Analytics Builder Workflow Canvas showing dataset transformations with filters selects renames and joins.

In the example above, the Organizational Units dataset branches into parallel paths for:

Sections
Offerings
Programs
Semesters
Departments

This makes it easier to build hierarchies such as program → semester → section or other multi-level organizational structures.

Divergence is only supported in SPICE. DirectQuery can't support multi-branch workflows.

Composite datasets

Watch this video to learn how to use composite datasets in Analytics Builders.

After you build an organizational structure dataset, you can add it to any report you create. You do this by using the custom dataset as a data source for a new dataset. Composite datasets in Analytics Builder allow you to reuse custom data across multiple reports, save time, and ensure consistent results.

Switch between experiences

Watch this video to learn how to switch between experiences in Analytics Builder.

Datasets created in the legacy data preparation experience can never be converted to the new experience because they use a different internal structure that is not compatible with the new visual workflow editor. Authors must rebuild their legacy datasets manually in the new experience.

To convert a dataset from the New Data Preparation experience to the legacy experience

Create a new dataset. It automatically opens in the New Data Preparation experience.
Click the Switch back to legacy experience opt-out link in the top right of the interface, below the toolbar.
Your new dataset reopens in the legacy data preparation workflow.

You may switch back and forth between experiences until the dataset is saved. After you save or publish the dataset in the new experience:

The opt-out link disappears.
All future edits open in the new experience.
The dataset can't be converted back to the legacy interface.

Analytics Builder search results for 'UserLogins' including 5 search results. The legacy dataset is shaded darker and doesn't have (newprep) after its name.

If you prefer to continue using the legacy experience, you can create a new dataset and start over, or duplicate an existing dataset that was originally created in the legacy editor.

FAQ

Why are some datasets greyed out when I try to add them in the New Data Preparation experience?

Datasets can only be combined when they were created in the same experience. This means:

Legacy parent datasets can only be used by child datasets also built in the legacy experience.
New Data Preparation experience datasets can only be used together with other New Data Preparation experience datasets.