AWS lambda & selenium & python is powerful solution, but requirements conflicts are sensitive. You can upload them as lambda layer and you don't need to be careful and upload them every time ;)

• run publish_selenium_layer.sh
• open lambda function and use scraping.py as example.
• make sure you extend lambda itself time out. (The default is 3 seconds)
• import the layer above script created.

That's it!

This selenium binaries works with python 3.7. NOT with 3.8 :(

If you want Python 3.8, you need to deploy as docker container image instead of lambda layer, pleate visit docker-selenium-lambda

If you don't want to create function and import this layer for each scraping purposes, please visit my project pythonista-chromeless