Postdoctoral Fellow – Agentic AI Decision Support for Neurodegenerative Disease Diagnosis

APPLY

Please send a cover letter describing your interests, relevant experience, and specific fit for our lab and this project, a CV, and three names of references to Pedro.Pinheiro-Chagas@ucsf.edu.

Job Summary

The NeuroAI Lab at UCSF, led by Professor Pedro Pinheiro-Chagas, is seeking applications for a postdoctoral fellow to advance the Memory and Aging Center (MAC) Co-pilot: an agentic, AI-powered decision support system for neurodegenerative disease diagnosis. The successful applicant will join a multidisciplinary team combining expertise in artificial intelligence, clinical neurology, neuropsychology, and radiology, aiming to create next-generation diagnostic tools that affect real-world clinical practice. The primary project will extend the MAC Co-pilot, an agentic AI system that integrates large language models, discriminative/predictice machine learning, and clinician-facing interfaces to synthesize multimodal data—including clinical notes, neuropsychology, and neuroimaging—for accurate neurodegenerative disease diagnosis. In collaboration with world-class specialists Bruce L. Miller, Katherine Possin, Maria Luisa Gorno-Tempini, and Bill Seeley, M.D. from the UCSF MAC, and Andreas Rauschecker from the Radiology Department, this highly collaborative initiative offers comprehensive mentorship and direct translational impact. Supported by the UCSF Catalyst Award, the MAC Co-pilot project is not only focused on academic innovation but is also positioned for commercial translation with the potential to become a widely deployed clinical product. The appointed fellow will spearhead the development, evaluation, and optimization of the MAC Co-pilot’s AI workflows, elaborate dynamic prompt engineering (e.g., RAG, knowledge graph integration), and implement robust evaluation protocols using real and simulated clinical data. Responsibilities include improving accuracy and interpretability, automating multimodal data processing, developing research and production-ready software, and maintaining full compliance with data security and regulatory standards (HIPAA) on the UCSF Versa AI platform.

Required Qualifications

• Ph.D. in Computer Science, Biomedical Engineering, Neuroscience, Data Science, or related quantitative discipline.
• Demonstrated expertise in machine learning, with strong foundations in deep learning and generative AI with Python.
• Experience with natural language processing and large language models, especially in retrieval-augmented generation or multi-agent frameworks.
• Proven record working with health-related or cognitive/clinical research datasets. (Experience with neuroimaging, clinical notes, or neurodegenerative disease data highly desirable.)
• Established track record of peer-reviewed publications and exceptional scientific writing skills.

Preferred Qualifications

• Experience with (or strong interest in) agentic AI architectures, workflow orchestration, and human-AI interaction design.
• Knowledge of data security, privacy, and clinical regulatory requirements (e.g., HIPAA compliance) for clinical software.
• Interest or experience in entrepreneurship, technology translation, or the commercialization of research innovations.
• Demonstrated capacity to work autonomously and thrive in multidisciplinary research teams, adapting effectively in fast-paced, translational environments.