Building models isn't enough

Hello geeks, I think that building a model using either machine learning or deep learning isn’t enough. Deploying the model for production is a crucial step that needs attention from beginners like me. Are there any recommended courses or books?


Beginners need to understand the mechanics of models first and foremost!


Yes, this is very important.

Unfortunately, this is extremely complicated because there are so many different possible platforms and user cases.

1 Like

@BALK I’m currently going through Chip Huyen’s Designing Machine Learning Systems [Book] and am finding it pretty good. While it doesn’t discuss so much exact specifics for, as TMosh said there are many possible platforms out there, it does go through a lot of the ‘practical’ considerations around design and deployment-- Issues that are less thoroughly (if at all) covered in academic ML/DL classes (ex. how clean is your data ?), that might otherwise not be apparent as issues for a new comer, but are very important in production.

*P.s. I believe Chip is one of Prof. Ng’s former students and taught a course on this subject at Stanford.


If you are speaking to the operations side, you can search MLOps and do some reading.

The extent you need to know this probably depends on the company/job, imo. When assessing whether or not to learn something, I’ll skim job postings and see which technologies/platforms employers are seeking.