Start with simple files/tasks, cut in inexpensive/scrap material and build on successes and learn from failures.
For “real” projects, apply the “rule of 3” and plan on cutting it three times — first a test in scrap, then a dry run in some inexpensive material, then only commit and full-send for real once everything has been proved out.
Don’t underestimate the difficulties of workholding, and consider a belt-and-suspenders approach where both blue painters tape and mechanical workholding are used.
For 3D work, you may want to look into 2-sided operations:
or even making your own fixtures:
Note that there are entire books on workholding:
(from: https://www.goodreads.com/review/list/21394355-william-adams?ref=nav_mybooks&shelf=cnc)
and we have a fair number of references, esp.