Traditional n-tier multi-user applications should not be novel to anyone these days; but a significant portion of such applications is dedicated to addressing infrastructure issues such as scalability and reliability. But what if we looked at these applications as an platoon of independent, stateful microservices each encoding a state machine? In this talk, John shows how declaratively describing state-machines formalizes the behaviour of a message-based application and allows us to reason about its correctness - and introduces a way of expressing such state-machines as full-fledged Orleans Actors, with particular emphasis on type-safety, developer productivity, testability and ease-of use. No prior knowledge of State Machines is assumed, but a general idea of what Orleans is will be helpful.