Data science is a team sport. The conversation needs to include different stakeholders with different views. The usual people are other data scientists and business stakeholders.
If you’re in a Jupyter notebook, getting feedback from these people is possible but involves degrees of separation from the original notebook. It’d be much better to collaborate in the notebook itself, which you can do with other tools.
Getting feedback and comments in a Jupyter notebook
Comments from data scientists involve a code review workflow on platforms like GitHub. This is problematic for the same reasons that using Git for Jupyter notebooks is. It’s a workflow designed for software engineers.
Comments from business stakeholders involve taking parts of the notebook out and putting it into a convenient format, like screenshots in a presentation or an Excel export. This is removed from the notebook itself. When you make changes based on those inevitable comments, you’ll need to go through making sure your screenshots and exports are up to date.
Using a notebook with comments built-in
You can avoid all of these problems by using a Jupyter-compatible notebook tool such as Deepnote that lets you make comments directly inside the notebook. As they’re close to the code, data scientists can leave comments easily. As the user interface is approachable for everyone, business stakeholders can leave comments as well.