IBM Developer Blog

Follow the latest happenings with IBM Developer and stay in the know.

Learn about our experience with the Jupyter ecosystem and see some of the contributions IBM has made during the past year.

For this year’s JupyterCon, which is held online October 5-17, we are excited to have the opportunity to share our experience with the Jupyter ecosystem and showcase some of the contributions IBM has made during the past year.

Note that the time specified in the tutorial and general session links is UTC/GMT.

Tutorial sessions (October 5-9, 2020)

Tutorial sessions, which are 90-minute, instructor-led, hands-on activities, are held October 5-9, 2020.

Developing extensions for JupyterLab

Introduced in JupyterLab v2 in 2019, extensions can be used to customize and extend any part of JupyterLab. For example, an extension might add a new file editor, add Git versioning support, or apply a theme. The Developing Extensions for JupyterLab tutorial on October 9 covers the basics and provides you with an opportunity to build a simple extension.

Clustering algorithms using Python and scikit-learn

If you want to learn about the theory and ideas behind unsupervised learning, attend the Clustering algorithms using Python and scikit-learn tutorial on October 6.

Open source fundamentals

The Open source fundamentals sessions on October 9 discuss key ideas of open source development, such as licenses, governance, funding, and community.

General sessions (October 12-16, 2020)

General sessions, which are 15- to 30-minute talks, are held October 12-16, 2020.

Some of our JupyterCon sessions cover various aspects related to JupyterLab extensions, such as how to use, create, or debug them. Other sessions are dedicated to Elyra, a set of AI-centric extensions to JupyterLab, and its most prominent feature — the Visual Pipeline Editor. Using this editor, you can assemble AI pipelines from notebooks or Python scripts and run them in JupyterLab or Kubeflows Pipelines, a popular platform for building and deploying portable, scalable machine learning workflows based on Docker containers.

So whether you are new to JupyterLab or an experienced user, there’s likely something interesting for you in our sessions:

Debugging notebooks and Python scripts in JupyterLab

If you are a Jupyter Notebook user reaching out to external IDEs to debug your code and haven’t had the opportunity to familiarize yourself with the JupyterLab Debugger extension, you should consider adding Karla’s talk in the Jupyter Community: Practices track on October 13 to your agenda. In her session, she will discuss how to set it up and demonstrate the main debugger features.

A generic metadata-store for JupyterLab extensions

When we created Elyra, we needed a generic way to store metadata and a user-friendly approach to managing it in the JupyterLab user interface. In their talk in the Jupyter Community: Tools track on October 14, Martha explains how we implemented a generic metadata service extension front end and back end, and demonstrates how they can be utilized by other JupyterLab extensions, such as a file editor.

Introduction to Elyra: an AI-centric extension for JupyterLab

Do you have a set of notebooks (or Python scripts) that download data, process and analyze data, train a model, or perform any other machine learning workflow tasks? With the introduction of the Visual Pipeline Editor in Elyra, you can now run them as part of a pipeline to automate repetitive steps. Join Yiwen, Edward, and Saishruthi in their session on October 12 in the Jupyter Community: Tools track to learn how to use the Elyra Visual Pipeline Editor to create workflow pipelines and how to run them in JupyterLab or a Kubeflow Pipelines deployment. The demo scenario is based on a set of notebooks that process a popular time series data set from the Data Asset eXchange.

What is new on Elyra: a set of AI-centric JupyterLab extensions

Luciano and Karla provide an overview of Elyra and demonstrate the main features in their talk on October 12 in the Jupyter Community: Tools track. Covering Git integration, the code snippet editor, Python editor, Visual Pipelines Editor, and much more, this session is ideal if you are looking for ways to simplify common tasks in JupyterLab.

Explore and extend AI pipeline runtimes with Elyra and JupyterLab

Elyra’s Visual Pipeline Editor makes it easy to assemble machine learning workflows from Jupyter Notebooks or Python scripts. Alan’s deep dive on October 12 in the Jupyter Community: Tools track provides interesting insights into how Elyra processes the pipeline with the help of a platform like Kubeflow Pipelines.

Building AI pipelines with Elyra: a deep dive using COVID-19 analytics scenario

In their talk on October 15 in the Jupyter Community: Practices track, Luciano, Fred, and Karla demonstrate how data scientists on our team used Elyra’s Visual Pipeline Editor to create and run a Jupyter Notebook-based machine learning pipeline that downloads, transforms, cleanses, and analyzes COVID-19 time series data. You can find the notebooks and pipeline referenced in this talk in the covid-notebooks GitHub repository.

IBM Quantum Experience notebooks. Serving JupyterHub at scale for the quantum computing community

Focusing on enterprise adoption of the Jupyter ecosystem, Juan covers in his session in the Enterprise Jupyter Infrastructure track on October 14 how the IBM Quantum Experience uses a JupyterHub deployment on Kubernetes to serve notebooks to the quantum computing community, which has more than 200,000 users. No quantum computing experience is required to enjoy this talk!

Can’t make it to JupyterCon this year?

If you can’t attend the conference this year but would like to learn more about Elyra and try it out, we’ve got you covered:

  • The overview summarizes the main features of Elyra.
  • The installation guide has details on how to install Elyra using pip, conda, and from source. You can also run Elyra in a Docker container, or, if you don’t want to install anything at all, in the cloud at
  • If running your notebooks or Python scripts as a pipeline in JupyterLab or on Kubeflow Pipelines is something you would like to try, take a look at the tutorials.

Would you like to get in touch with us before, during, or after the conference?

Elyra is a community-driven open source project that is maintained by a small number of developers and data scientists. If you’d like to file an enhancement request, report an issue, or simply ask a question, head over to