Insights agent

View how to ask questions about your data in plain English, follow up on answers and see where each number comes from.

How the insights agent works in Drivetrain

Everything builds on a question you ask in plain English. From there, check where the numbers come from, follow up on the answer and lay it out as a table.

1
Ask a question in plain English
- Type a question about your data the way you would ask a colleague, for example active headcount by department for a given month.
- Drive AI looks across the metrics, dimensions, time periods and versions in your Drivetrain workspace to answer it.
- Get a structured answer with a breakdown and key insights, without building a report.
2
Trace each answer to its metric, dataset and version
- Each answer lists the sources behind its numbers.
- See which metric, dataset and version the numbers came from.
- Open a source metric to see how it is calculated.
3
Ask a follow-up question that keeps the context
- Follow up in the same conversation, for example to compare with the same month last year and show the net change.
- Drive AI keeps the context, so there is no need to repeat filters or start over.
- See the overall change, the change for each group and key takeaways.
4
Show the answer as a table
- Ask for the same results as a table.
- Compare groups side by side, spot changes and check the numbers at a glance.

How Drivetrain simplifies ad hoc analysis

Answer the questions leaders and budget owners send you without building a new report each time.
Follow up on any answer, such as a comparison with last year, without rebuilding filters or versions.
See which metric, dataset and version every answer comes from, so you can check the numbers.

FAQs

Can Drivetrain answer ad hoc questions in plain English?

Yes. You type a question the way you would ask a colleague. Drive AI looks across the metrics, dimensions, time periods and versions in your Drivetrain workspace and returns a structured answer, without building a report.

Can Drivetrain show which data an answer is based on, so I can verify it?

Yes. Each answer lists its sources, so you can see which metric, dataset and version the numbers came from and open the metric to see how it is calculated.

Can Drivetrain answer follow-up questions?

Yes. Drivetrain keeps the context of the conversation, so a follow-up such as a comparison with the same month last year builds on the previous answer.

Can Drivetrain show an answer as a table?

Yes. Ask for the results as a table and Drivetrain lays out the same numbers in rows and columns, which makes it easier to compare groups and check the figures.