Deepnote, a 'Successor to Jupyter Notebook', Goes Open Source
SAN DIEGO — Deepnote, an analytics and data science notebook for teams, has recently made its platform open source (https://github.com/deepnote/deepnote).
Since its inception in 2019, the company has experienced tremendous adoption, with over 500,000 data professionals from some of the best data teams in the world using Deepnote as their primary notebook.
In a keynote address at JupyterCon, founder and CEO of Deepnote, Jakub Jurových, expressed his pride in the company's achievements.
"We built a notebook that was easy to use, beautiful, and we aimed to do more. We wanted to enhance our users' experience and contribute more to the community," Jurových said.
Deepnote, built on Jupyter notebook, has long aimed to address the challenges users have found with Jupyter, according to Jurových.
"We tackled issues like lack of native integrations, messy and confusing UI, and the fear of beginners and non-technical users," he explained.
Stability, versioning, and the problem of 'works on my machine' were also areas Deepnote identified for improvement in data science notebooks, Jurových told the conference audience.
Collaboration was another problem Deepnote sought to solve. In a blog post, Jurových described Deepnote as the 'successor to Jupyter notebook'.
"We're opening the format and building blocks to provide a standard purpose-built for AI," he wrote, emphasizing the shift from single-player JSON scrolls to reactive, AI-ready projects.
What Does Deepnote Offer?
The open-source Deepnote includes several notable features:
- AI agents, with single-player authoring 'coming soon'. Deepnote's agent assists in writing, editing, and explaining code.
- A shared workspace for collaboration between technical and non-technical users, accommodating native versioning, comments, reviews, and human-readable projects.
- Blocks for various functionalities, including SQL queries, Python/R code blocks, charts, tables, inputs, file upload, buttons, layout, and reusable modules.
- Data app creation capabilities, allowing users to deploy notebooks as interactive dashboards or data apps with a single click.
- Over 100 native integrations with governed secrets, enabling quick and secure connections to data sources without the need for password copy-pasting.
- Reactive execution, ensuring automatic updates to downstream blocks, solving reproducibility issues.
- Compatibility with Jupyter, VS Code, Cursor, and Windsurf.
Jurových highlighted the key benefit of data science notebooks, emphasizing their ability to bring technical and non-technical users together, making them a powerful and accessible computational medium.
"Notebooks are set to define the next decade of computing," he concluded.
