At Databricks, we purpose to make it easy for enterprises to harness knowledge to hurry up enterprise processes and improve decision-making. AI/BI is our AI-first enterprise intelligence product, natively built-in into the Databricks Platform. It’s out there now to all prospects of Databricks SQL.
AI/BI Dashboards are a strong technique to securely share key enterprise metrics, visualizations, and insights throughout your group. They are often shared with different customers in your Databricks workspace, with customers in different workspaces in your group, in addition to with Databricks view-only customers (customers who’ve been added to the Databricks account, however to not any workspace). On this publish, we’ll take a look at among the key AI/BI Dashboards options that allow dashboard sharing throughout your group.
Making a shareable dashboard
Let’s begin with an instance. We constructed an instance Dashboard that highlights gross sales alternatives and we need to share it with our gross sales group. To make it partaking, we have included quite a lot of visualizations, which is able to interactively filter one another, permitting the gross sales group to zero in on the info that issues to them. Moreover, to make sure the dashboard performs easily, we restricted every dataset to not more than 100K rows. This permits Databricks to load all the info straight into the consumer’s browser, leading to lightning-fast dashboard loading and interactions.
In case you are creating your first AI/BI Dashboard, take a look at this tutorial to get began. To be taught extra about find out how to make your dashboards extra interactive, check out our documentation on filters, parameters, and cross-filtering. And for efficiency optimization, right here’s some steering on find out how to make your dashboards quick.
Publishing the dashboard
Throughout the dashboard improvement course of, we collaborated with varied co-authors on the draft model of the dashboard. Now that it’s able to share extra broadly, we are able to Publish the dashboard. Publishing creates a snapshot of the present configuration, so when it’s shared, viewers will see this polished model as an alternative of any ongoing edits. We are able to at all times republish the dashboard to replace the snapshot later.
When publishing an AI/BI Dashboard, you may select whether or not or to not embed credentials. In the event you embed credentials, viewers will entry the dashboard’s knowledge and SQL warehouse utilizing your credentials, which is useful in the event that they don’t have direct entry to the supply knowledge or SQL warehouses. With out credential embedding, viewers will use their very own credentials, which is critical if you wish to personalize content material based mostly on the consumer or preserve strict knowledge entry controls (e.g. as declared in Unity Catalog).
In my case, my gross sales group are view-only customers. They’ve been added to our Databricks account, permitting them to view dashboards, however haven’t been added to any workspace. Thus, they don’t have entry to the dashboard’s SQL warehouse, so we’re embedding credentials to make sure they’ll work together with our dashboard. When embedding credentials, It’s necessary to make sure that the shared knowledge complies together with your group’s knowledge entry insurance policies.
Sharing the revealed dashboard
Now that the dashboard is revealed, it’s time to share it with the gross sales group. We are able to share with any customers or teams which are a part of our group’s Databricks account. They don’t must have entry to our workspace.
There are two major methods to share the dashboard with account customers:
- Sharing with particular person customers or teams: You will discover them within the sharing dialog’s search outcomes.
- Sharing with the entire group: You possibly can set the dashboard to “Anybody in my group can view.”
On this situation, we selected to share with the Gross sales Staff group, which my Databricks admin has synced with our identification supplier. (In case you are an account admin, the next part consists of extra info on find out how to add view-only customers to your Databricks account.)
If we’re sharing with a small group (20 folks or much less), they’ll mechanically obtain an e-mail notification from Databricks. For bigger teams, like our gross sales group, or when sharing to the entire group, no automated e-mail is shipped, so we’ll must manually share the hyperlink. To do that we’ll copy the hyperlink from the button on the backside of the Share dialog and place it on our firm wiki.
Since our gross sales group are view-only customers and don’t have entry to our Databricks workspace, they’ll be prompted to log in by way of our identification supplier after they click on the shared hyperlink. As soon as logged in, they’ll see a dashboard-only expertise. They won’t have entry to every other Databricks workspace content material.
Including view-only customers to the Databricks account (for account admins)
Databricks account admins have the flexibility so as to add view-only customers. That is carried out by including customers to the Databricks account, however not granting them permission to entry any workspace. Customers with out workspace entry are restricted to viewing shared AI/BI Dashboards however they’ll’t edit them or entry different workspace content material like Workflows, ML Experiments, and many others.
Admins can do that in a number of methods:
- Manually add customers or teams to the Databricks account and easily don’t grant them workspace entry to maintain them view-only. (docs)
- Use the Databricks Person APIs to programmatically handle consumer entry. (docs)
- Use SCIM to sync customers and teams from Entra IDOkta, or one other identification supplier. (docs)
- Straight use Seamless sharing for Entra ID customers and teams in Databricks (presently in non-public preview, speak together with your account consultant for entry).
- Simply-in-time account consumer provisioning (presently in non-public preview, speak together with your account consultant for entry), which mechanically provides customers to the Databricks account after they log in by way of your identification supplier.
Conclusion
AI/BI Dashboards in Databricks supply a fantastic alternative to share important insights and drive data-driven decision-making throughout your group. Whether or not you’re sharing with a choose group or your total enterprise, the pliability and energy of AI/BI make it simple to create interactive, quick, and safe dashboards that everybody can profit from.
We’re excited to see the way you and your groups will leverage these capabilities to remodel what you are promoting processes. We encourage you to dive in and begin creating your personal dashboards in the present day. To get began, discover our AI/BI Dashboards documentation and comply with our tutorial to construct your first dashboard.
For many who need to see these options in motion, take a look at our on-demand session from DAIS 2024 on YouTube, the place we showcase the newest in AI/BI, knowledge integration, and dashboard sharing. These classes are full of insights that can provide help to maximize the worth of Databricks AI/BI inside your group.
Do not hesitate to succeed in out to our group in case you have any questions or want steering in your journey. We won’t wait to see the modern methods you employ AI/BI Dashboards to empower what you are promoting and drive significant outcomes.
Glad dashboarding!
