Speech AI technologies such as automatic speech recognition (ASR), text-to-speech (TTS), and neural machine translation (NMT) automate millions of conversations today. They provide customers with a personalized human-like experience through such applications as Virtual Assistant, Contact Center Agent Assist, and digital avatar. We'll review a few inspiring use cases and solutions to common challenges faced by enterprises that are getting started with speech AI. Learn techniques for achieving world-class accuracy and customizing for your industry. We'll also show the latest models, tools, and features in NVIDIA NeMo and Riva, demonstrate sample applications, and illustrate how to build a speech AI pipeline.