Research Assistant Professor, Applied AI: Agent Evaluation & Data Governance

Role Summary

Eshelman Innovation, housed in the nation's #1 pharmacy school, was created to address translational challenges with a game-changing new approach to academic innovation and commercialization.

Its venture studio runs on Kalphai, an agentic AI venture builder spanning the full venture lifecycle, from intake and evidence-grounded assessment of new concepts through venture coaching, program management, and an AI-assisted path toward pilot-ready software, operating today on real faculty deal flow under the University's security and data-governance controls and architected in partnership with AWS.

Eshelman Innovation seeks a dynamic leader to establish the AI assurance and data governance program that makes Kalphai deployable, defensible, and commercially viable across academic medical centers and university innovation organizations.

The chosen candidate will serve as the Principal Investigator for this program and will focus on producing the evidence, tooling, and documentation that allow a partner institution to entrust its innovation portfolio to the platform.

This position provides an opportunity to define what trustworthy agentic AI looks like in a real institutional deployment, rather than in a benchmark.

Minimum Education and Experience Requirements

Ph.D. in computer science, machine learning or artificial intelligence, biomedical or health informatics, data science, statistics, human-computer interaction, software engineering, or a related field. Candidates whose doctoral training lies outside these fields but who bring directly relevant technical research experience are encouraged to apply.

Preferred Qualifications, Competencies, and Experience

Research or applied experience evaluating machine-learning or AI systems, ideally including large language models or multi-agent systems, with a measurement-first stance toward any claim about how a system behaves. Strong programming ability in Python, TypeScript, or comparable languages, and demonstrated comfort building research tooling against a production codebase, across services, cloud infrastructure, and version-controlled deployment. Working understanding of data-governance and privacy principles including de-identification, provenance, retention and deletion, access control, and tenant isolation.

Experience red-teaming or safety-testing LLM applications, or building evaluation and observability tooling for agentic systems. Familiarity with health-data regulation and governance, including HIPAA, FERPA, IRB processes, and technology transfer or Bayh-Dole context, or experience in academic medical, university innovation, or other regulated settings. Experience with retrieval-augmented generation, agent frameworks, or cloud AI services such as AWS Bedrock and AgentCore, and with infrastructure-as-code and multi-tenant architectures. A record of translating technical work into compliance-facing, procurement-facing, or otherwise non-technical documentation. Experience working with security or privacy reviewers, or with responsible disclosure practice.

Experience leading or contributing to complex technical collaborations, including mentoring students or junior colleagues. Excellent written and verbal communication, including publishing in respected refereed venues and presenting to varied academic, institutional, and commercial audiences.