Carol Willing

Project Jupyter, evolved from IPython and involving over a thousand contributors, is an open-source platform for interactive computing which has become a de facto standard for data science in research, education, journalism and industry. With millions of users and a novel, extensible architecture, Project Jupyter has a deep impact in interactive computing. It continues to grow as an open and inclusive collaboration between academia, government, philanthropic funders and industry, led by teams at UC Berkeley, Lawrence Berkeley National Laboratory and Cal Poly.

Project Jupyter exists to develop open source software, open standards, and services for interactive computing across dozens of programming languages. These enable users in scenarios where human-driven exploration meets computation and data, a class of problems that is now pervasive across all sectors of society. These diverse use cases are supported by its extensible architecture, which separates the execution context (kernel) from user interfaces. While the project team has developed the IPython kernel and the Jupyter Notebook, with features optimized for interactive computing workflow in Python, its architecture is language-agnostic. Third parties have developed alternative user interfaces, extensions, and kernels for over 100 programming languages and environments, ranging from C++ to database systems.

Today more than 2,000,000 Jupyter notebooks are on GitHub, each a distinct instance of a Jupyter application, covering technical documentation to course materials, books and academic publications. Jupyter has been transformative in scientific collaborations and reproducibility, as exemplified by its use at the LIGO observatory, whose discovery of gravitational waves was recognized with the 2017 Nobel Prize in Physics: the LIGO Open Science Center publishes Jupyter Notebooks that allow anyone to replicate their original analyses. JupyterHub supports the deployment of Jupyter tools in multiuser environments, from small research groups to universities, companies and other organizations. JupyterHub is used in numerous commercial companies, research at facilities such as CERN and high-performance computing centers like NERSC and SDSC. In education, Jupyter has been used to author "executable textbooks" in signal processing, statistics, machine learning, bioinformatics, data science and more. JupyterHub allows institutions to deploy entire live courses: UC Berkeley's education initiative in data science, data8.org, offers students an interactive textbook in the cloud that only requires a web browser and university credentials. Similar setups exist at multiple educational institutions, adapted to their specific needs. Jupyter has gained wide industry adoption. Since 2015, Jupyter-based products have been released by Google (Cloud DataLab), Microsoft (AzureML, HDInsight), Intel (Trusted Analytics Platform), IBM (Data Science Experience) and Kaggle, to name a few. These are commercially hosted environments based on Jupyter, with company-specific additions. O'Reilly uses the Jupyter architecture (kernels to notebooks) in its media publishing pipeline, including books, websites and the Oriole interactive system. Bloomberg and Anaconda Inc have partnered with the Project to develop its next-generation user environment, JupyterLab.