topic Re: Databricks AI/BI Dashboard cross-filtering across different dataset in Warehousing & Analytics

Databricks AI/BI Dashboard cross-filtering across different dataset

lennox1 — Tue, 16 Sep 2025 19:14:59 GMT

Hello Everyone,

I recently started using AI/BI dashboard and I'm finding it difficult to cross-filter two visuals from two different dataset. Visual 1 displays Sales by region (month to date) sourced from dataset 1 and Visual 2 displays Open orders by region (month to date) sourced from dataset 2. Is there a way when I select South region on visual 1, visual 2 also only displays information for South region?

Re: Databricks AI/BI Dashboard cross-filtering across different dataset

BS_THE_ANALYST — Tue, 16 Sep 2025 19:53:48 GMT

@lennox1 I've just gave this a try and didn't have much luck for cross-filter across multiple datasets. The documentation seems to indicate the same thing: https://learn.microsoft.com/en-us/azure/databricks/dashboards/#cross-filter

I scanned the release notes for 2024 and 2025 with CTRL+F. There's a good deal of updates for cross-filtering. Could be worth keeping your eye out: https://learn.microsoft.com/en-us/azure/databricks/ai-bi/release-notes/2025

To filter different datasets together, you could use a Global filter? ☺️. I appreciate it's not quite cross-filtering but it's a cool thing nevertheless. I.e.

Before filter

After global filter

Some other community members may have some funky workarounds.

All the best,
BS

Re: Databricks AI/BI Dashboard cross-filtering across different dataset

JoeMo — Mon, 09 Feb 2026 14:18:36 GMT

We are finding the same issue that you cannot cross filter on multiple datasets. This would be an incredibly useful feature. Our scenario is we have a trip with a set of deliveries. We want to find inefficient trips based on trip level KPIs and then display the details of the deliveries once this trip is selected. Keeping trips and deliveries into 2 separate datasets to hopefully increase speed.

However, it seems like i have to build a single dataset of trips and deliveries and then deal with trip level KPIs aggregation to make sure something like average miles per trip does not count a trip multiple times because there is a row in the data for each deliver on that trip.

Cross-filtering with multiple datasets would make any 1 to many relationship much easier to deal with in the dashboards.