Clinical Medicine Domain Expert

What you'll do

The physician counterpart to the Pharmaceutical R&D Domain Expert role, and structured the same way: a full-time W-2 job with Cincinnatus LLC, the employer of record, placed inside a leading AI lab's GenAI team as extended workforce. You work in the lab's own tools, on client-issued equipment, next to its research and program management teams.

The brief is to improve how frontier models reason about real clinical work:

• Data QA and review: checking clinical tasks and model outputs for missing behaviours, thin reasoning, unsafe or off-guideline recommendations, and answers that read well but would not hold up on a ward round
• Instruction specs and golden datasets: writing specs and reference solutions to clinical problems, and defining new tasks that reflect how medicine is actually practised
• Benchmarks: designing difficult clinical tasks and evaluation sets, and building medicine-specific skills and tools with the researchers
• Calibration: agreeing standards with researchers and specialists in adjacent fields, and turning tacit clinical judgment into criteria others can apply

For a clinician, the realistic comparison is a full-time move out of practice into an AI lab, with a six-month initial term. That is a larger decision than taking on a few hours of remote evaluation, and the listing should be weighed that way.

Who fits

• MD or DO from an accredited school, with a completed residency in a recognised specialty
• 4+ years of post-residency clinical practice; residency and fellowship years do not count
• Active, unrestricted licence in at least one US state, plus board certification in your specialty
• A defined specialty (internal medicine, oncology, radiology, emergency medicine, surgery, psychiatry or a subspecialty are named). Utilization management, clinical informatics or medical affairs experience is a plus
• Senior standing: attending, medical director, division chief, associate or full professor, CMO
• Everyday professional use of LLMs, and the judgment to tell a sound answer from a plausible wrong one
• 40 hours a week, reliably, for an initial 6 months
• Precise, well-structured written feedback

Location

Hybrid, Bay Area, California. You must live there and be on-site with the client's team several days a week when required, and relocate at your own cost if you are not local. No relocation assistance. Not a remote role.

What it pays

$70–110 per hour, $5 below the pharma sibling at both ends. At 40 hours that annualises to about $146,000–229,000 before tax, which is below typical attending compensation in most specialties. The trade is a W-2 job with benefits administered by Cincinnatus and direct involvement in model development. Weekly Stripe or Wise payment terms are not published on this listing.

Worth knowing

Good:

• W-2 employment with payroll and benefits, not a contractor arrangement
• Direct work with an AI lab's research team rather than through a task queue
• A route into clinical AI for physicians considering a career change
• Informatics, UM and medical affairs experience counts in your favour

Less good:

• Pay is well below most attending salaries
• Full-time and hybrid: you cannot keep a normal clinical schedule alongside it
• Relocation at your own cost if you are outside the Bay Area
• Initial 6-month term with nothing said about what comes after
• The lab is not named, and the on-site pattern is "when required"

About this listing

Found on Mercor; employment, onboarding, payroll and benefits run through Cincinnatus LLC, a separate legal entity. Confirmed open on 24 September 2026. The pay band, 40-hour commitment, 6-month term, licensure and Bay Area requirement are the ad's own. See Mercor.