Data Science Workspaces

A dream IDE for data science and model development.

We spent a year perfecting our ideal data science IDE, building on the Eclipse Foundation's excellent Theia project. The result is a cloud IDE that is purpose built for ML model development, data science, and data visualization, without any compromises. 100% in the browser. No downloads or installation required. Data Science Workspaces are an instant data science environment for your VPC.

Jupyter notebooks.

Our team has been using Jupyter notebooks for over a decade. The creator of Jupyter was Plotly's first advisor, and Plotly was the first corporate sponsor of nteract. We're champions for the Jupyter notebook products and community.

With Data Science Workspaces, you can bring Jupyter notebooks into your corporate environment through Dash Enterprises's secure, scalable Kubernetes platform.

To create Dash apps within Jupyter notebooks (e.g. classic Notebook, JupyterLab, Visual Studio Code notebooks, nteract, PyCharm notebooks, etc.), please see Jupyter Dash.

Dash app authoring & deployment.

Once you've developed a model in Jupyter notebook, you can build a Dash UI for the model and deploy in to Dash Enterprise, all without leaving your Data Science Workspace. Once deployed, the rest of your company can access and run the model through your Dash app UI.

The Dash app on the left dynamically bins 42 million records using Python's Vaex library.

Enterprise grade.

Data Science Workspaces bring Python, R, & Julia scripting and notebooks to the enterprise environment through corporate authentication, a Secrets Manager, Kubernetes scaling, turnkey HPC, and easy installation on Azure, GCP, or AWS.

Docker.

Each Data Science Workspace runs in its own isolated Docker container for optimum performance, security, and customizability. Because Dash Enterprise uses Kubernetes, managing and scaling these Docker containers is automatic. The Docker architecture allows each Workspace to have its own notebooks, scripts, data files, and package dependencies without effecting other Workspaces or Dash apps running on Dash Enterprise.

Terminal.

Dash Science Workspaces have a fully functional terminal where you can install new Python or R packages, run shell scripts, and launch Dash apps.

In this screen recording, the user adds Tensorflow to their Workspaces environment through the built-in terminal, then launches a Dash app.

Typeahead.

Dash Science Workspaces support first-rate autocompletion, so you can spend less time referencing documentation and debugging typos.

Plugins.

Data Science workspaces already ships with dozens of plugins to optimize your development experience, and we're adding more each month. Plugins already enabled include:

• Changing the IDE color theme
• Launching a Dash app
• Shortcuts for Terminal commands
• Viewing CSV or Excel files in an editable table

Most VSCode plugins will work with Data Science Workspaces. Get in touch if there is a plugin that you'd like to see!

Secure networking

Workspaces run on servers in your corporate network enabling data scientists access to network-secured data stores and APIs. Run multiple Dash Enterprise platforms in different networks to give finer grained control to networked data stores without compromising security and without expanding network access to individual laptops.

Linux is King 👑

For data science and ML, Linux is King. Your work laptop may be Windows, but your Data Science Workspace on Dash Enterprise is Linux within an isolated Docker container.

Data science packages, database drivers, and production-grade Dash apps run best on Linux. Data Science Workspaces run in secure, isolated Linux-based containers that is accessible in an IDE in the web browser. This means that your data scientists can continue to run Windows on their workstation while writing code on a Linux operating system via their browser.

With deployed Dash Applications running the same operating system as Workspaces, you'll spend less time reconciling the differences between production and development.

Dev = Prod.

Every developer has struggled with reconciling the differences between a development environment and their production environment. Subtle differences between networking, operating systems, Python packages, database drivers, virtual environments, CPU & memory and more cause a myriad of difficult-to-debug issues when bringing your application into production. Further, many data scientists's development environments differ from one another causing friction when sharing apps and projects among a team.

In Data Science Workspaces, the Docker-based development environment is run on the same network and inside identical Docker containers as deployed applications. All of the Dash Enterprise features like Job Queues, Onboard Redis and Postgres Databases, PDF Reporting, and Low-Code packages are available. You'll never run into issues where an app works in development but fails in production.