I would prototype this by duplicating a vertical slice of the business that is (presumably) well understood and easy to teach. Don't start with the hardest thing and definitely don't try to plan it all up front. Hire some cheaper contract labor to begin with. If you find you can't even replicate a small part of something you thought you understood, you should probably abandon the overall effort.
The last thing I would do is make some grand decision, locate an 8-figure budget, hire a bunch of people, and then hope for the best. Incrementalism is the key to not blowing this idea up. You already have a ~working system. Leverage the hell out of that. At the end of the day this is probably going to be more of a people problem than a technology problem - managing scope and expectations throughout, etc.
The last thing I would do is make some grand decision, locate an 8-figure budget, hire a bunch of people, and then hope for the best. Incrementalism is the key to not blowing this idea up. You already have a ~working system. Leverage the hell out of that. At the end of the day this is probably going to be more of a people problem than a technology problem - managing scope and expectations throughout, etc.