JetBrains Unlocks Team Development with On-Prem Data Science Notebook Environment

(iluistrator/Shutterstock)

Data science teams that want to work together in a the Jupter notebook environment but don’t want to run in the cloud may be interested in the new Datalore for Enterprise offering unveiled last week by JetBrains.

Datalore combines the power of the Jupyter data science notebook with PyCharm, JetBrain’s integrated development environment (IDE) for Python. This gives users smart coding features from the PyCharm IDE, such as auto-complete, documentation assistance, and code refactoring.

The software gives users access to data science capabilities that they’re used to from the Jupyter notebook–such as exploring and cleaning data, creating machine learning and deep learning models, creating interactive visualizations, and sharing the results–but without the need to set up the notebook environment, according to JetBrains.

Datalore enables users to organize team tasks in shared workspaces. They can also collaborate on Jupyter notebooks in real-time, and create history checkpoints for version management. Static copies of the notebook can also be shared via a link.

Previously, JetBrains offered two versions of Datalore: the community edition, which is free, and Datalore Professional, which adds a certain number of CPU and GPU hours. Last week, the company launched a third version, Datalore Enterprise, which adds new collaboration and runtime capabilities.

Datalore Enterprise enables teams of data science users to collaborate in a Jupyter-style environment in an on-prem manner

With Datalore Enterprise, data science teams can now collaborate in the same Jupyter environment while running the entire system on-premise. That was a new feature that JetBrains customers will be excited to get, says Alena Guzharina, Datalore’s product marketing manager.

“Over the past years, we received lots of in-bound requests for an on-premises solution and team plans as well,” Guzharina says in a press release. “Previously, you could have run open-source solutions on your own infrastructure, but their installation and customization took significant efforts and time. Datalore for Enterprise makes this process effective and introduces a smart team ecosystem for real-time collaboration, and we’re proud to announce its launch.”

Datalore for Enterprise runs on customers’ Kubernetes clusters, or installs on customers’ private AWS clouds. Subscriptions to the software cost $125 per month per person. The Prauge, Czech Republic-based company is giving customers a month of free access to the product and a dedicate support team to test the tool. For more information, see www.jetbrains.com.