Ryan Louie

News

✨ I am on the job market in Fall 2025! I am currently looking for full-time research roles in the US academic and industry market in the areas of Human-Centered AI and LLMs for Mental Health ✨.

About me

I am a postdoc in Stanford's Computer Science department, where I am part of the Stanford NLP and Stanford HCI groups and advised by Diyi Yang and Emma Brunskill. I received my PhD from Northwestern's Technology and Social Behavior program, where I worked closely with Prof. Haoqi Zhang and Darren Gergle on Human-AI Interaction and Social Computing research topics.

My research advances human-centered AI approaches to create systems that assist in activities for promoting personal well-being. I have built human-AI systems that assist with training novice counselors to effectively support patients (mental health), empower casual creators to compose music that expressively communicates their ideas or emotions (creativity), enabling friends to connect across distance (social connection). Motivated by these application domains, I invent technical methods for effectively customizing and interacting with AI capabilities (LLMs, generative models, context-aware AI) so that system behaviors are aligned with the goals and values of their designers and end-users. Then, I rigorously evaluate the systems via experimental studies and real-world deployments to develop a scientific understanding of how these human-AI systems impact human behavior and user experience.

My current agenda develops and evaluates scalable, human-centered AI systems to train the next generation of mental health counselors. My collaborators and I have developed CARE, an LLM-based training system that enables novice counselors to practice with AI-simulated patients and receive feedback from AI mentors. To achieve this, I tackle technical challenges in aligning LLMs with domain expertise and address empirical gaps in evaluating AI training systems. My work uses experimental methods—including randomized trials with 90+ counselors and micro-randomized trials in psychotherapy classrooms—to understand how LLM-simulated practice can upskill and enhance educational activities.