[Remote] Healthcare AI Trainer - 257964

Note: The job is a remote job and is open to candidates in USA. Medix is seeking experienced clinical practitioners to evaluate frontier AI models on realistic clinical work. The role involves designing clinical scenarios, assessing and comparing model outputs against medical standards, developing grading rubrics, documenting failure modes, and reviewing tasks created by other clinicians.

Responsibilities

• Design Realistic Clinical Tasks: Construct authentic clinical scenarios based on day-to-day practice. Author supporting clinical files (constructed or fully de-identified patient histories, exam findings, lab results, imaging reports, medication lists, and prior notes) and craft precise prompts framing the task as you would brief a trusted colleague
• Evaluate AI Model Outputs: Run scenarios through frontier AI models and grade deliverables (assessments, plans, reconciliations, summaries, protocols) against the standard you would hold a professional colleague to
• Side-by-Side Model Comparison: Compare outputs from two AI models on identical prompts and files, determine which performed better, and document precise clinical deficiencies
• Develop Granular Grading Rubrics: Write detailed rubrics specifying what a correct deliverable must contain—such as correct differential diagnoses, red-flag identification, medication safety checks, and appropriate tone—and explain why a response passes or fails
• Identify & Flag Failure Modes: Provide evidence-based documentation for clinical errors, including unsupported assessments, missed drug interactions or contraindications, diagnostic anchoring bias, hallucinated/ignored data, unsafe recommendations, or off-brief interpretations
• Peer Review & Collaboration: Review, refine, and stress-test task sets authored by other clinicians across various medical specialties

Skills

• You hold an active or previous clinical credential (e.g., MD, DO, NP, PA, PharmD, RN) with substantial hands-on experience in patient care, clinical documentation, and complex medical decision-making
• You possess strong diagnostic instincts, sharp attention to detail regarding drug safety/reimbursement/protocol compliance, and an eye for subtle clinical oversights
• You can translate complex medical reasoning into clear, structured, written feedback and rubrics
• You are comfortable working with digital tools, building structured datasets/scenarios, and working independently in a remote tech environment
• You are eager to find the limits of current AI models by crafting tasks that a competent professional would handle correctly, but that a model currently fails to navigate

Benefits

• Contract position for 12 weeks
• 10+ hours/week with a flexible schedule—you create your own schedule
• 100% remote; work from anywhere
• Impactful work helping build safer, more reliable, and clinically sound AI tools

Company Overview