Principal Researcher

Job Summary

The Principal Researcher provides a supportive research environment to a highly qualified recent PhD. The Principal Researcher collaborates with Emma Levine and Pietro Veronesi to enable research for the CLARITY (Clinician-Led AI Resources Individualized to You) project. The PR will be a major collaborator on scholarly work and may be required to present that research at academic conferences. This position is expected to last one year with the possibility of a second.

• Conduct collaborative research with faculty to be presented at academic forums and published in peer‑reviewed academic journals.
• Collect data and conduct statistical analysis.
• Conduct literature reviews and perform research.
• Lead system architecture decisions, model selection, and fine‑tuning strategies for the CLARITY AI pipeline.
• Design and implement end‑to‑end integration of LLM‑based text processing with video generation tools.
• Develop guardrails to ensure factual accuracy, appropriate uncertainty communication, and clinical reliability.
• Lead the development of a secure, patient‑facing interface that allows patients to access, review, and audit their personalized videos.
• Collaborate with other staff and faculty on the CLARITY project.
• Design experiments to test prompting strategies and personalization features; analyze experimental data.
• Support clinical trial development alongside medical faculty.
• Use subject matter and best practices knowledge to perform laboratory and/or research‑related duties and tasks; work independently to assist with project design.
• Guide data collection, evaluation, and analysis. Design and manage databases. Write reports, articles, and manuscripts.
• Perform other related work as needed.

• ABD status by start date, PhD preferred by start date in a field relevant to the research being conducted, such as computer science, artificial intelligence, machine learning, computational linguistics, or a closely related field. PhD required after one year of employment.

• Knowledge and skills developed through 5‑7 years of work experience in a related job discipline.

• Demonstrated experience working with large language models (LLMs) and/or multimodal generative AI systems.
• Designing, implementing, and evaluating applied AI systems in real‑world settings.
• Fine‑tuning or adapting foundation models.
• Background with human‑centered AI, AI alignment, or safety/guardrail design.
• Building user‑facing AI systems or web‑based interfaces.
• Working with healthcare data or other high‑stakes domains.

• Demonstrated ability to lead collaborative technical teams.
• Outstanding verbal, written, and presentation skills, as well as organizational skills.
• Work independently with little supervision; possess a self‑motivated disposition; identify opportunities for improvement and recommend effective changes, all while achieving key objectives resulting in desired outcomes.
• Excellent strategic planning, critical thinking, analytical, and persuasion skills.
• Handle multiple detailed tasks/projects simultaneously and meet strict deadlines with frequent interruptions.
• Demonstrated ability to work effectively and diplomatically with colleagues, as well as with students, faculty and corporate contacts in a multitude of communication methods, including in person, email, and phone.
• Professional demeanor, including tact, discretion, and a customer‑service‑oriented approach.

• This position is currently expected to work a minimum three days per week in the office.

The University of Chicago offers a wide range of benefits programs and resources for eligible employees, including health, retirement, and paid time off. Information about the benefit offerings can be found in the Benefits Guidebook.

The University of Chicago is an equal opportunity employer and does not discriminate on the basis of race, color, religion, sex, sexual orientation, gender, gender identity, or expression, national or ethnic origin, shared ancestry, age, status as an individual with a disability, military or veteran status, genetic information, or…