This document discusses concurrency, scalability, and fault tolerance using actors and software transactional memory (STM) in Akka. It describes how actors provide message-passing concurrency by isolating processes and avoiding shared state. Actors allow easy reasoning about concurrency by preventing issues like race conditions and deadlocks. The document also covers how Akka implements actors, typed actors, remote actors, supervisor hierarchies for fault tolerance, and STM for coordinating concurrent changes to shared state transactionally. Combining actors with STM results in "transactors" that allow building distributed, concurrent, and fault-tolerant applications.