About the Role
We are deploying generative AI agents across clinical operations and global development workflows: clinical data review, anomaly detection in trial data, protocol Q&A for site teams, Informed Consent Form generation and translation, and protocol element document assembly.

Each system needs rigorous evaluation. In specialized clinical contexts, "good" is not a single number. It depends on the user, the regulatory context, the therapeutic area, and the cost profile of errors. We are looking for a postdoctoral researcher to help us develop the evaluation frameworks, rubrics, and quality systems that determine when these AI agents are ready for our scientists, clinical operations specialists, and site teams to rely on.

Key Responsibilities
What you'll do

• Develop evaluation frameworks and rubrics for AI agents operating in clinical operations and global development contexts
• Define quality dimensions for specialized outputs, e.g. clinical data queries, anomaly detection, protocol Q&A responses, ICF content, regulatory documents, that reflect how clinical-ops users actually judge quality
• Design and run human evaluation protocols with clinical, regulatory, and site-facing subject matter experts
• Build benchmark datasets and evaluation pipelines that the broader GenAI organization uses across AI applications
• Partner with our AI engineering teams to translate evaluation findings into measurable improvement
• Contribute to the methodology of how a regulated pharma organization evaluates AI in clinical work

What we're looking for
Domain expertise

• Strong understanding of clinical operations and global development workflows, e.g. clinical data management, monitoring, regulatory documentation, site operations, or related areas
• Familiarity with how clinical-ops specialists, monitors, regulatory authors, and site teams actually do their work and assess quality
• PhD or equivalent research experience in a clinical, biomedical, regulatory, biostatistics, or closely related field

Enthusiasm for AI agents

• Genuine interest in working hands-on with generative AI and agentic systems
• Curiosity about how LLMs behave in specialized domains and where they fail
• Comfort prototyping with modern AI tooling (LLM APIs, vector databases, evaluation frameworks)

Aptitude for evaluation systems

• Demonstrated ability to develop frameworks, rubrics, and structured assessment protocols
• Experience designing evaluations with multiple quality dimensions, expert raters, or non-trivial inter-rater reliability considerations
• Strong analytical and writing skills; comfort with ambiguity and unsolved problems

Practical skills

• Python and the modern ML/data stack
• Comfort working across disciplines, e.g. clinical operations, regulatory affairs, AI engineering, data science

Practical details

• Duration: 2-year postdoctoral appointment
• Reports to: [TBD]

Works closely with: Generative AI Teams, Scientific Fellow, clinical operations and regulatory subject matter experts across J&J global development

At Johnson & Johnson, we believe health is everything. Our strength in healthcare innovation empowers us to build a world where complex diseases are prevented, treated, and cured, where treatments are smarter and less invasive, and solutions are personal. Through our expertise in Innovative Medicine and MedTech, we are uniquely positioned to innovate across the full spectrum of healthcare solutions today to deliver the breakthroughs of tomorrow, and profoundly impact health for humanity.