You need to build a bunch of inexpensive, low-effort prototypes. Do that while you're employed. The most important thing here is what will the startup _do_, no

You need to build a bunch of inexpensive, low-effort prototypes. Do that while you're employed. The most important thing here is what will the startup _do_, not how to run it. Spend some time figuring that out. Anything you spend time on that ISN'T the product (renting out houses, consulting jobs) are a waste of time.