Yeah, but then you'd have to work in consulting. ;)
A related idea is to just take the 400 bucks and spend it on AWS playing around. Spin up a few Kubernetes clusters, try out their various database offerings, maybe see if you can run a load-test and get your play system to auto-scale.
That would also require keeping a sharp eye on the billing side to make sure the total cost doesn't go over $400... but, hey, that's part of the learning, too!