Daniel Arnst works as a Senior Conversational AI Architect for Telefonica Germany and is responsible for standardizing and automating the digital voice assistant and chatbot Aura, as well as analyzing customer behavior. With his linguistic and phonetic background, he brings machines closer to understanding human language. Before joining his current position, Daniel worked at Amazon Alexa, where he focused on speech components such as text-to-speech, automatic speech recognition (ASR), and natural language understanding (NLU). He was responsible for quality assurance of rule-based and machine learning NLP models and contributed to building Alexa's lexicon. As a computational linguist, he has also been involved in software development and has been working on NLU design and best practices for model training for several years. Currently, he is exploring the opportunities and risks of large language models for conversational AI.