cancel
Showing results forย 
Search instead forย 
Did you mean:ย 
Get Started Discussions
Start your journey with Databricks by joining discussions on getting started guides, tutorials, and introductory topics. Connect with beginners and experts alike to kickstart your Databricks experience.
cancel
Showing results forย 
Search instead forย 
Did you mean:ย 

Building Plug-ins for AI/BI Dashboard

suryaprayaga
New Contributor

I want to develop some plug-ins in Databricks to allow me build 3D charts and visualisations that I can use code stubs from external sources like ObservableHQ. The visualisations can be written in any framework or language, but I was looking for a way where I can achieve this within Databricks.

The thought came from building a Grafana dashboard that has rich visualisations and admirable aesthetics.

Any thoughts?

suryaprayaga
1 ACCEPTED SOLUTION

Accepted Solutions

AmitDECopilot
Contributor

My understanding is Databricks AI/BI dashboards are not really built like Grafana where you can install custom visualization plug-ins inside the dashboard itself. So if the requirement is true custom 3D charts using ObservableHQ / Three.js / custom JS, I donโ€™t think AI/BI dashboard will directly support that as an internal plug-in model today.

What I would consider instead:

  1. Use Databricks for the data prep / SQL / model output layer.
  2. Build the custom visualization separately using Observable, Plotly/Dash, Streamlit, React, or another web framework.
  3. Host it through Databricks Apps or another internal app hosting option if available in your environment.
  4. Keep AI/BI dashboards for standard BI/reporting use cases, and use a custom app when the visualization experience needs to be very rich or interactive.

So basically, Databricks can still be the backend/data platform, but for Grafana-like custom plugin style visualizations, I would separate the visualization layer instead of trying to force everything inside AI/BI dashboard.

Amit Kumar Singh
Lead Data Engineer | AI-Assisted Data Engineering

View solution in original post

1 REPLY 1

AmitDECopilot
Contributor

My understanding is Databricks AI/BI dashboards are not really built like Grafana where you can install custom visualization plug-ins inside the dashboard itself. So if the requirement is true custom 3D charts using ObservableHQ / Three.js / custom JS, I donโ€™t think AI/BI dashboard will directly support that as an internal plug-in model today.

What I would consider instead:

  1. Use Databricks for the data prep / SQL / model output layer.
  2. Build the custom visualization separately using Observable, Plotly/Dash, Streamlit, React, or another web framework.
  3. Host it through Databricks Apps or another internal app hosting option if available in your environment.
  4. Keep AI/BI dashboards for standard BI/reporting use cases, and use a custom app when the visualization experience needs to be very rich or interactive.

So basically, Databricks can still be the backend/data platform, but for Grafana-like custom plugin style visualizations, I would separate the visualization layer instead of trying to force everything inside AI/BI dashboard.

Amit Kumar Singh
Lead Data Engineer | AI-Assisted Data Engineering