Skip to main content

The Basics of Building an AI/BI Dashboard in Databricks

Summary

  • Dashboard in minutes: Build a production-ready AI/BI dashboard with multi-page layouts, filters, and visualizations in minutes using Databricks.
  • AI-powered visualizations: Use the Databricks Assistant to automatically create charts and visualizations, then customize them as needed for your analysis.
  • Genie integration: Enable Genie to ask natural-language questions directly on the dashboard and get instant data insights without writing SQL.

The Basics of Building an AI/BI Dashboard in Databricks

Watch: The Basics of Building an AI/BI Dashboard in Databricks
🎯 What You'll Learn:
How to quickly create a production-ready AI/BI dashboard using Databricks
How to create and manage data sources registered in Unity Catalog
How to build a multi-page dashboard
How to add and utilize basic filters
How to create visualizations quickly with the Databricks Assistant
How to enable Genie feature to get data insights by asking questions directly on the dashboard
How to publish a dashboard
How to use global filters
How to embed the dashboard into other tools
⏱️ CHAPTERS:

Chapters

FAQs

How quickly can you build a production-ready dashboard in Databricks?

This video demonstrates building a complete AI/BI dashboard in a matter of minutes, including data sources, multi-page layouts, filters, visualizations, and publishing without requiring extensive development time.

What is the Databricks Assistant in dashboards?

The Databricks Assistant uses AI to automatically create visualizations and charts based on natural-language requests, making it easy to build analytics dashboards without writing code or SQL queries.

How does Genie work in dashboards?

Genie is a feature that lets you ask natural-language questions directly on the dashboard about your data, providing instant insights and automatically generating the underlying SQL queries for transparency.

Can you embed Databricks dashboards into other tools?

Yes, Databricks dashboards can be published and embedded into other applications and tools, allowing you to integrate your analytics into external workflows and platforms seamlessly.

Full transcript

0:00 Introduction to the time it takes to build a production-ready dashboard

How long does it take to build a production ready dashboard? Is it hours, days, maybe even weeks? What if I told you that we could do it in a matter of minutes? You could create seni diagrams, bar graphs, geographic maps, and even use AI. I'm a technical marketing engineer here at Data Bricks. And in this video, I'll show you how you can build an interactive AIBI dashboard. Let's dive in. I am in the data bricks workspace homepage and first we'll navigate to the dashboard section to create a dashboard.

0:30 Creating and managing data sources

I need to create data sources from SQL. So I'll add code that combines different tables employee data with performance and compensation. I'll save that data source. Next, I'll add the termination data table as a second source that is already registered in Unity catalog. Now, we're ready to build the executive summary page. I'll add a text box at the top with markdown for the title. I'll name this dashboard HR I'll save the global filters for later.

1:03 Building a multipage dashboard

We'll come back to those in a bit. Moving on to summary metrics. First up is the total number of employees. I'll use a count from the termination data. I'm using the data bricks assistant to make it super easy to do this. The count looks good, so I'll accept, resize, and move on. Next, we'll look at average salary from the employee data. I'll format this as a dollar amount so it's easy to read. I'll add another measure and format it as a dollar amount, too.
I can resize these and drag and drop to arrange them however I like. I'll also add some basic filters to help us explore the data. I'll start by including a filter for employees based on the state that they live in. I'll add another filter that filters by the region that employees live in. Lastly, I'll add a filter based on the employees hired date. This will filter everything downstream, which will help when getting insights around retention.

2:24 Adding and demonstrating basic filters

Check out the relative date filtering options we have available. Last year, 60 days, last 12 months, since last year. There's a lot of flexibility here. Or if you prefer, you can select specific months, days, and years. Let's create some visualizations. We can use the data bricks assistant to do this. I'll ask for hires by department. And I want to show this by year. Even though we use the data bricks assistant, we can still manually go in and make edits.

3:30 Creating visualizations using the Databricks Assistant

For example, I'd like to group by department. Notice how I can select the grain daily, monthly, or annually. I'll go with annually for this one and accept this visualization. All right, my dashboard is done. Now I can go ahead and enable Genie. This will allow me to get data insights just by asking questions on my dashboard. Time to publish. First, I'll set the data set permissions. I can either share my permissions or let the individual use their own.

4:18 Publishing the dashboard

I can also share the dashboard with people or groups of users. And here's my published dashboard. Here you can see I have my filters. I have some counter visualizations, um, bar graphs, line graphs, sink keys, scatter plots, and at the very bottom here, I even have a forecast. I have another page of data, which is geographic analysis. I have a point map and I have a table. My global filters work across all pages. As you can see, everything is filtered to the HR department.
So, I can see information across all pages without having to change the individual filter. I am also going to embed this dashboard into Confluence so that other business users more familiar with that experience can access the dashboard there. In Confluence, it's just a matter of selecting an iframe, adding the URL, and making sure the width is appropriate so that everyone can see your new dashboard. From blank canvas to interactive dashboard, AI suggested visualizations when we needed them.

5:13 05:49: Embedding the dashboard into other tools

We customized exactly what we needed from a dashboard. And what's the result? A production ready dashboard with advanced analytics right here on Data Bricks. >>

Learn more about the Databricks Data and AI platform.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.