Analyze your data

ACL AI Studio offers an interactive approach to explore datasets and unlock meaningful analysis using conversational AI and collaborative tools.

Analyses are conducted interactively; you edit, run, and iterate on queries, with AI assistance available throughout. Each query, result, and conversation is saved as part of an analysis library linked to your dataset. You can create output tables (read-only snapshots) with AI-suggested names, ready for sharing or export.

Generate analysis

Note

This procedure includes steps for both the new UI and old UI. Select the appropriate tab (New UI or Old UI) in this section based on your interface.
You can use the old UI until June 30, 2026. Starting July 1, 2026, the new UI will be the only available experience.

Steps

On the ACL AI Studio home page, from the list of datasets, select the dataset for which you want to analyze data.

On the Dataset summary page, ACL AI Studio automatically generates AI-powered summaries highlighting key trends, statistics, and notable data points.
From the list of AI-suggested analysis ideas, select the one that you want to get more insights.

The analysis idea is organized by category. The selected idea is displayed in the conversational workspace.

To get more options, select Refresh.

If you want to customize the suggestions, from the drop down menu next to Refresh, select Customize suggestions.
Modify the idea, if required, and select (the start analysis icon) to generate the analysis.

A page appears with two sections:
- The query section on the left with a Reasoning area where you can view the natural language and SQL query used to generate the output table. It shows how the data was filtered, transformed, or calculated, helping reviewers understand the logic behind the results.
- The section on the right displays the interim output table.
In the query section, select the pencil icon to further enhance the table using natural language or SQL query, and select Run.
Select Save.
In the Save output dialog box, assign a custom name or use the AI-suggested name and select Save.
You can view the saved table in the Output tables section of Tables.

In an ongoing analysis, you can analyze an existing source or output table.

From the Tables list, select the source or output table that you want to analyze.

The table opens beside the conversational workspace.
Select Analyze this table.

Edit an analysis

Only Owners and Editors assigned to a dataset can edit an analysis. All analytic modifications and outputs are linked to the dataset’s lineage and audit trail, ensuring compliance and reproducibility.

From the list of output tables, select the three dots next to the analysis that you want to edit.
Select Edit.
On the Edit analysis panel, update the analysis title and description as per your need.
Select Save.

Delete an analysis

Deleting an analysis in ACL AI Studio is a fully governed, role-based action that maintains the integrity of your audit and lineage records, supporting transparency and compliance throughout the platform. Owners and Editors assigned to the dataset have permission to delete analyses.

From the list of output tables, select the three dots next to the analysis that you want to delete.
Select Delete.
In the confirmation dialog box, select Delete.

Steps

On the Dataset home page, from the list of datasets, go to the dataset for which you want to analyze data, and select View details.

On the details page, ACL AI Studio automatically generates AI-powered summaries highlighting key trends, statistics, and notable data points.
Select Start analysis.
On the New analysis page, from the list of AI-suggested analysis ideas, select the one that you want to get more insights.

The analysis idea is organized by category. The selected idea is displayed in the conversational workspace.

Note

To get more options, you can select the + Get more button displayed under a category.

If you dismiss a suggestion, it doesn't appear again for the analysis.
Modify the idea, if required, and select the run icon () to generate the analysis.

The generated analysis is saved under the Analysis library section of the dataset.

Output tables

Output tables are generated at the end of each completed analysis, capturing the filtered, transformed, or computed results as read-only data.

Tables are visible and reusable for subsequent analyses. Any output table can act as a starting point for new queries or further data exploration.
Only final output tables are retained; interim tables generated during analysis are not saved.

To save an output table, follow these steps:

On the workspace page of the dataset, go to the Analysis library section and select the Open analysis button next to the analysis that you want to work with.
On the analysis details page, select the expand icon () next to the table to view the entire table.
On the page that appears, use natural language or SQL query to further enhance the table.

Note

The query section records the natural language or SQL query used to generate the output table. It shows how the data was filtered, transformed, or calculated, helping reviewers understand the logic behind the results.

To hide the query section, select Hide query.
Select Save.
In the Save output dialog box, assign a custom name or use the AI-suggested name and select Save.

Edit an analysis

On the workspace page of the dataset, go to the Analysis library section and select the three dots at top right corner of the analysis that you want to edit.
Select Edit.
On the Edit analysis panel, update the analysis title and description as per your need.
Select Save.

Delete an analysis

On the workspace page of the dataset, go to the Analysis library section and select the three dots at top right corner of the analysis that you want to delete.
Select Delete.
In the confirmation dialog box, select Delete.