How do you build services for high volume, fault tolerance, reliability, and asynchronous design patterns?