The human voice captures what surveys and clicks cannot: Emotion in real-time, cognitive load under pressure, personality through prosody, and authentic reactions before a more conscious reflection. Yet, most behavioral researchers lack the tools to leverage this rich data source. In this 5-day workshop, you’ll master the complete pipeline from audio processing and feature extraction to developing deployable conversational agents for primary data collection. You’ll learn to extract acoustic features that reveal distinct psychological states, transcribe and analyze natural language data, and design voice-based AI systems that can conduct interviews or to develop experimental interventions at scale.
What makes this course unique:
•Theory meets practice: We cover the fundamentals of speech science and behavioral research foundations while building production-ready interfaces and reproducible analytics pipelines.
•Full-stack skills: You will lean the foundations from audio preprocessing to feature extraction to LLM-powered conversational agents all in one course.
•Hands-on from day one: We analyze real voice data, build working prototypes, and you will leave with a mini-project tailored to your own research.
•Open and responsible: We also cover ethical data practices and how to share your methods transparently with others.
This course is ideal for PhD students and researchers in marketing, psychology, economics, communication, HCI, and related fields who want to expand their methodological toolkit and apply voice analytics methods and develop conversational AI agents in their research.
