SAN DIEGO — Deepnote, an analytics and data science notebook for teams, is now open source.The company has seen tremendous adoption of its platform since it started in 2019.“Over 500,000 data professionals from some of the best data teams in the world have made Deepnote their primary notebook, and we were pretty proud of what we built,” said Jakub Jurových, founder and CEO of Deepnote, in a keynote address at JupyterCon here.“We built a nice notebook that was easy to use. It was beautiful, but we also felt like we could do more. We could do more for our users, and we could also do more for the community.”Deepnote, built on Jupyter notebook, has long aimed to address some of the challenges users have found with Jupyter, according to Jurových.“We tried to solve these problems one by one, problems such as the lack of native integrations, problems with the UI, which was messy and confusing. It was pretty much scary to all these beginners and all the nontechnical users.”Stability, versioning and “also the old problem of ‘works of my machine,’” were also areas where Deepnote saw room for improvement in data science notebooks, Jurových told the conference audience.Collaboration was also a problem that Deepnote sought to solve, the CEO said. In a post on his company blog in which he called Deepnote the “successor to Jupyter notebook,” he expanded on this idea.“We’re also doing this because the center of gravity has moved: from single‑player JSON scrolls to reactive, AI‑ready projects that humans and agents can co‑author, review, and deploy,” he wrote. “We’re opening the format and the building blocks so the community has a standard that is purpose-built for AI.”What Does Deepnote Do?So what’s in open source Deepnote? Among the features Jurových called attention to:AI agents, with single-player authoring “coming soon,” according to the company. Deepnote’s agent “helps to write, edit and explain the calls,” Jurových said in his keynote.A shared workspace where technical and nontechnical users can collaborate, accommodating native versioning, comments, reviews and “human‑readable projects with clean diffs,” according to Jurových’s blog post.Blocks for, as his blog post lists them, “SQL queries, Python/R code blocks, charts, tables, inputs (text/number/select/slider/date), file upload, buttons/actions, layout/app pages, reusable modules.”Data app creation capabilities, such as for users to deploy notebooks as interactive dashboards or data apps in a single click.More than 100 native integrations with governed secrets, “so you can connect your data sources quickly and securely,” Jurových said in his keynote. “We no longer need you to copy-paste passwords and secret tokens in your notebooks.”Reactive execution, so downstream blocks update automatically. These reactive kernels “solve all of those problems with reproducibility,” Jurových told the keynote audience.Compatibility with Jupyter, VS Code, Cursor or Windsurf.One of the key benefits of data science notebooks, Jurových said in his keynote, is that they are “among the very few computational mediums that actually allow you to bring technical and nontechnical users together. It’s a place that’s easy to get started with, but also can be very powerful.”He added, “We believe that notebooks are a computational medium that is going to define the next decade of computing.”The post Deepnote, a ‘Successor to Jupyter Notebook,’ Goes Open Source appeared first on The New Stack.