Confession of a New Course-aholic
Learnings and Upcoming Events
Hi all!
Thank you for joining me in this beautiful sun from Seattle and reading my newsletter. I have been a little bit aloof in what is happening with my Jupyter Notebook to Production course that should have been out in November. First, thank you all for being patient and allowing me to explain why there is such a huge delay.
Teaching Philosophy
I write a lot about simplicity and clean code, which also translates into my course philosophy. The goal is to share complex ideas in simpler terms and make them accessible to everyone (especially those with terrible math teachers).
For the courses, this means that I will place a high emphasis on smaller sections, shorter videos, code snippets, and learning.
I know that grading will help the students learn more and better (read Make It Stick: The Science of Successful Learning). As I progress through this journey, I will have to play around with quizzes and assignments. Bare with me in this adventure, as I want to ensure you get the best education possible.
Course Progression
For my first course, I wanted to teach something I had seen other folks teach, but it never stuck with me. I wanted to relate to my students as much as I wanted to get a better grasp of the material. Machine learning satisfied both desires and became my first course, “Machine Learning 101”.
My second course, on the other hand, was strategic. A common theme I discovered was that data folks with various ranges of experience struggled to get out of the Jupyter Notebook and into production. The second course is about how to get yourself from getting yourself out of a jupyter notebook to production, and the title will be “Jupyter Notebook to Production” (or at least something along those lines).
Two major obstacles arose as I progressed through building the second course. The first was setting up code snippets. These were not as important in the first machine learning course as most concepts are visualized using graphs and tables. I was using Kajabi for my courses, and though the platform is awesome, it is lacking in the code department. I did not realize how bad it would be until I got into this course.
The second major roadblock was me. Wait, Sarah, how can you be an obstacle? I got myself too deep too quickly. I realized too far into the second course that I have much more to teach than I originally planned, so I need to split the last course into two courses. The first course will be “How to data project” and the second will be “Productionalizing your data projects”. The titles are not finalized, but those are the general ideas.
How to data project will cover the basics of data projects (ha!). First, it will cover the differences between popular roles and how your data project will differ if you are a data analyst versus a data engineer or scientist. Second, I want to understand how a data project will look for each role.
Productionalizing your data projects will cover how a data scientist, data engineer, or machine learning engineer can get their jupyter notebook or python files ready for production. This will cover things like linting, writing clean code, setting up classes (and best practices), and then we will cover how to monitor data pipelines. I could add machine learning pipelines, but if I am being realistic, that would also be an entire class of its own.
All of these changes will take some time, but overall, it will be a better experience for you. That is what matters to me. I have to translate the previous code into a new platform that is more code-formatter friendly. The second thing is creating courses just takes a lot of time, so they may not be out till January 2023. I will let you know when I have a better idea of when what comes out.
Final thoughts
I have many thoughts about how I want to grow these courses and the platform itself. And so I will have to go because I have my work cut out! Thank you for reading, and if you enjoy this newsletter, share it with your friends.