Our lab aims to to translate the dynamic, moment-by-moment analysis of psychotherapy into actionable clinical tools. We view Therapist Responsiveness, the ability to adapt actions to the patient's continuously evolving state, not just as a trait, but as a complex process of decision-making that can be modeled and improved (Bugatti & Boswell, 2016; Kramer, 2025). To support this, we are developing computational models that formalize the feedback loop between Context (the patient's in-session evolving state), Intervention (the therapist's action), and Micro-Outcome (the immediate effect). By integrating rich clinical data with advanced language models, we aim to create an AI-informed feedback system. This system will provide therapists with explainable insights and summaries of session dynamics, enhancing their clinical sensitivity, pedagogical learning, and real-time decision-making.

References:
Bugatti, M., & Boswell, J. F. (2016). Clinical errors as a lack of context responsiveness. Psychotherapy, 53(3), 262.

Kramer, U., Boehnke, J. R., & Esposito, G. (2025). Therapist responsiveness in psychotherapy: Introduction to the special section. Psychotherapy Research, 35(1), 1-3.‏