The most difficult part of my job as a data scientist is getting non-technical stakeholders to understand how data scientists do data science. yet another Data science solutions can help they Make better decisions.
But this is nothing new to me. This was also the case in my 5+ years of experience as a data scientist and machine learning engineer.
After much trial and error, the following order worked for me:
- Share regular progress updates (presentation slides) By simplifying technical concepts.
- Building machine learning web applications Towards the end of the project, we provide stakeholders with experience in working with the solutions we have built together.
However, the punchline was that a colleague of mine who had been on the same team for about five years was building a desktop application (rather than a web application) using .NET for a different use case. The team loves it.
So I asked myself. Why build a desktop instead of a web application?
However, there was one problem. Not only do I not know anything about .NET, but I've never built a desktop application before. Oops.
I knew that Python was a general-purpose programming language, so I wanted to see if it was possible to build desktop applications directly from Python.
After a few searches on Google, I found the following two frameworks:
Pyside6 and Tkinter both use Python as a wrapper to build desktop applications. Exactly what I was looking for. After reading the respective tutorials, I decided to try out PySide6. I'll probably try Tkinter someday, but that's not the point.
You'll think less and execute faster.
So I jumped right in.
Surprisingly, it wasn't as difficult as I thought.