With my wife, we are working on visual interface for creating Python scripts in the notebook. We created desktop application MLJAR Studio. In our app, user has a list of predefined steps. Each step has a graphical interface with a form that after filling generate the Python code. The Python code is the source of the truth.<p>Currently we have a few steps for training Machine Learning model on tabular data. Here you have few gifs with screenshots: <a href="https://mljar.com/docs/how-does-python-notebook-work/" rel="nofollow">https://mljar.com/docs/how-does-python-notebook-work/</a> how it looks like, and example how to build ML model: <a href="https://mljar.com/docs/create-first-notebook/" rel="nofollow">https://mljar.com/docs/create-first-notebook/</a> on tabular data. The created notebook is compatible with Jupiter notebook.<p>In the near future, we are planning to add notebook scheduling and more steps (probably with some dynamic manager for steps loading). We see MLJAR Studio as an alternative to visual programming environments which are node based. Because the Python code is the source of truth, it offers a great flexibility to define new steps or to add custom Python code.<p>The app is desktop based (it is using electron framework). It automatically installs Python 3.9 with miniconda and required packages. The installation is local, without change to the environment path. You can see installation instructions here: <a href="https://mljar.com/docs/install-notebook/" rel="nofollow">https://mljar.com/docs/install-notebook/</a>. The application is only for Windows. If you are interested in MacOS or Linux versions, please fill the form <a href="https://docs.google.com/forms/d/e/1FAIpQLSeB5-hA326sBg9fg-ppVeYXDYHbRWYoKO18p4MdnxFYknA7aw/viewform" rel="nofollow">https://docs.google.com/forms/d/e/1FAIpQLSeB5-hA326sBg9fg-pp...</a> and we will notify you when ready.<p>If you would like to try the app (currently Windows only), it can be downloaded from GitHub release page: <a href="https://github.com/mljar/studio/releases" rel="nofollow">https://github.com/mljar/studio/releases</a>