Ideas for 'Blog More' challenge #CNC2018

I know it's a bit past the deadline, but I decided to try and catch up with 'Blog More' challenge. I narrowed my ideas to following 5:

1. Few words about Data Classes in Python - In the next Python version (3.7), one of added features will be 'Data Classes' described in PEP 577. In my blogpost I would like to shortly describe what they are and why you might want to use them.
2. Getting started with Python apps on AWS Lambda - Recently, I'm learning more and more about "serverless" and figured that it could be useful to describe how you can build Python apps with Serverless framework and AWS Lambda.
3. CLI tools that I'm using on daily basis - There's a lot of awesome CLI tools that greatly simplify tasks I'm dealing with every day.
4. How to set up private npm registry on AWS Elastic Beanstalk - In this blog post I'd like to show how you can quickly set up your own private npm registry (https://github.com/verdaccio/verdaccio). Platform that I'll be using in my tutorial will be Elastic Beanstalk, but since I'll be using Docker, it can be easily adapted to other cloud providers or running it locally.
5. Introduction to AWS Machine Learning Application Services for Pythonistas - Recently, Machine Learning is rapidly gaining popularity and are being used for a variety of applications. What if you'd like to enchance your Python app with AI-capabilities, but don't have resources to develop it on your own ? AWS Machine Learning Application Services to the rescue! In my blog post (or series of posts), I will describe what services has AWS to offer when it comes to ML-driven APIs and I'll show how Pythonistas can leverage services like AWS Polly, AWS Rekognition or AWS Transcribe to add image analysis or text to speech conversion to their services.

I have a hard time picking the final 3, so any feedback would be really appreciated!

Thanks for reading!