Create Reusable Data Products

This section briefly covers some extended capabilities Dataiku DSS provides to coders and data scientists.

Dataiku’s code integration allows you to develop a broad range of custom components and to create reusable data products such as Dataiku applications, webapps, and plugins.

Dataiku Applications

Dataiku Applications are a kind of DSS customization that allows projects to be reused by colleagues who simply want to apply the existing project’s workflow to new data without understanding the project’s details.

Therefore, a data scientist can convert the project into a Dataiku application so that anyone who wants to use the application only has to create their own instance of the Dataiku application. These applications can take one of two forms:

  • Visual Applications that allow you to package a project with a GUI on top

  • Applications-as-Recipes that allow you to package part of a Flow into a recipe usable in the Flows of other projects.

Webapps

You can extend the visualization capabilities of Dataiku DSS with interactive webapps. Dataiku DSS allows you to write three kinds of webapps:

  • Standard webapps made at least of HTML, CSS, and Javascript code that you can write in the DSS webapp editor. DSS takes care of hosting your webapp and making it available to users with access to your data project.

  • Shiny webapps that use the Shiny R library.

  • Bokeh webapps that use the Bokeh Python library.

The Academy course on Webapps covers how to create interactive webapps in Dataiku DSS in detail.

Plugins

Plugins allow you to implement custom components that can be shared with others. Plugins can include components such as dataset connectors, notebook templates, recipes, processors, webapps, machine learning algorithms, and so on.

To develop a plugin, you program the backend using a language like Python or R. Then, you create the user interface by configuring parameters in .json files.

Plugins can be installed from the Dataiku Plugin store (in your Dataiku DSS instance) by uploading a Zip file or downloading from a Git repository.