What's new

Highlights of the latest Dataiku DSS releases.
More details in our release notes

Model Settings Reuse V 5.1.4 - June 2019

Dataiku provides several ways to reuse model settings. This allows you to create custom model settings to use as templates.

Schema Propagation V 5.1.4 - June 2019

When there are changes to the columns of a dataset, you want to propagate those changes to all downstream datasets. You can use the Schema Propagation tool to do this.

PyCharm Integration V 5.1.4 - June 2019

Though Jupyter notebooks are integrated into the Dataiku interface, many Python developers use PyCharm. Dataiku offers several integration points with PyCharm.

Subpopulation Analysis V 5.1.3 - April 2019

For regression and binary classification models trained in Python (e.g., scikit-learn, keras, custom models), Dataiku can compute and display subpopulation analyses. These can help you to assess if your model behaves identically across subpopulations; for example, for a bias or fairness study.

Partial Dependence V 5.1.2 - March 2019

For all models trained in Python (e.g., scikit-learn, keras, custom models), Dataiku can compute and display partial dependence plots. These plots can help you to understand the relationship between a feature and the target.

Overview V 5.1 - January 2019

Dataiku Version 5.1 offers a state-of-the-art experience for hardcore coder data scientists, and adds capabilities to help organisations govern their data initiatives while complying with regulations. See the following video for a quick introduction to 5.1!

Git Integration for Plugins V 5.1 - January 2019

The plugin editor now features full Git integration, allowing you to view the history of a plugin, revert changes, and to push and pull changes from a remote Git repository.

Learn more in our tutorials and in the reference documentation

Git Integration for Code Libraries V 5.1 - January 2019

In the library editor of each Dataiku project, you can now import code from external Git repositories.

Learn more in our tutorials and in the reference documentation

RStudio Integration V 5.1 - January 2019

Though Jupyter notebooks are integrated into the Dataiku interface, many R developers use RStudio. Dataiku 5.1 offers several integration points with RStudio.

Learn more in our tutorials and in the reference documentation

GDPR Capabilities V 5.1 - January 2019

A new plugin allows you to enforce a number of GDPR-related rules on projects.

Learn more in our tutorials and in the plugin page

Copy/Paste Prepare Recipe Steps V 5.1 - January 2019

You can now copy and paste preparation steps, either within a single preparation recipe or across preparation recipes, or even across DSS instances.

Learn more in the reference documentation

Show/Hide Subflows V 5.1 - January 2019

You can now hide parts of the Flow in order to improve the readability of very large flows. You can easily hide all parts of a flow upstream/downstream of a single node.

Learn more in the reference documentation

Copy Projects V 5.1 - January 2019

You can now easily duplicate a DSS project, optionally duplicating the content of some datasets.

Learn more in the reference documentation

Deep Learning V 5.0 - September 2018

Define a Deep Learning architecture using the Keras library to build a custom model in Dataiku’s Visual Machine Learning. You can then train, deploy, and score the model like any other model created and managed in Dataiku DSS.

Deep learning offers extremely flexible modeling of the relationships between a target and its input features, and is used in a variety of challenging applications, such as image processing, text analysis, and time series.

Learn more in general about deep learning in our Deep Learning Guidebook, and learn specifically about the Dataiku implementation in our tutorials and in the reference documentation

Enhanced Collaboration V 5.0 - September 2018

Dataiku DSS is all about making teams more efficient, so we want you to promote your work to others! In 5.0, you can:

Docker and Kubernetes Containerized Execution V 5.0 - September 2018

Share the load! Some processing tasks can be spun off of a DSS Design or Automation node into hosts powered by Docker or Kubernetes. This is fully compatible with cloud managed serverless Kubernetes stacks!

  • Python and R recipes
  • Plugin recipes
  • In-memory machine-learning
Please see Running in containers for more information.

And much more ... V 5.0 - September 2018

Dataiku DSS 5.0 also brings the following:

  • Resource management with Cgroups
  • Revamped homepage and global navigation within Dataiku DSS
  • Organize projects into folders
  • Automatically stop idele Jupyter notebooks
  • Train XGBoost models on GPU

Find all details in our release notes.


For older releases