AI/ML Research Scientist — Clinical AI

Research Remote (Global) Full-time $110K – $160K + Equity

ResearchEvalsRAGClinical AI

ABOUT THE ROLE

Push the frontier of clinical-grade medical AI. You'll design evals, run model comparisons, prototype RAG pipelines over PubMed and clinical guidelines, and shape which models Premedice ships. This isn't pure academic research — it's research that has to ship to patients. You'll own the eval suite, lead ablation studies on prompt design and tool routing, and publish what we learn so the field gets better with us.

WHAT YOU'LL DO

Responsibilities

Design and maintain the clinical eval suite — differential diagnosis accuracy, lab interpretation, drug interaction, citation faithfulness

Run structured model comparisons across frontier medical LLMs (Med-PaLM, MedGemma, GPT-5 medical, Claude medical, Gemini medical)

Prototype and benchmark RAG pipelines over PubMed, OpenFDA, RxNorm, DailyMed, ClinicalTrials.gov, and Europe PMC

Lead prompt-engineering and tool-routing research — when does a model need a tool vs. its own knowledge?

Publish internal research reports and, where appropriate, public papers / blog posts

Work with engineering to ship research wins into production without breaking the safety bar

Investigate failure modes — hallucinations, citation drift, reasoning errors on edge cases

Design human-in-the-loop studies with clinicians to validate model behavior on real cases

Stay current on clinical AI literature and surface what we should adopt, ignore, or push back on

WHAT WE'RE LOOKING FOR

Requirements

MS/PhD in ML, NLP, biomedical informatics, or equivalent industry research experience

Strong Python and PyTorch — you can implement evals and analyses from scratch

Published work in clinical AI, biomedical NLP, or a closely related field (workshop papers count)

Experience evaluating LLMs — you know the difference between a real eval and a vibe check

Familiar with medical knowledge sources: PubMed, OpenFDA, RxNorm, SNOMED, ICD, clinical guidelines

Comfortable working with clinicians — you can ask the right clinical questions and translate answers into research

Bonus: experience with EU AI Act conformity assessments, MDR clinical evaluation, or FDA SaMD submissions

AI-native workflow: you use tools like OpenCode or Claude Code to run experiments and analyze results

READY TO APPLY?

Interested in this role?

Send us your resume and a brief note about why you're the right person for this position. We review every application.

Apply via Email View All Roles