Role overview

Build the infrastructure that makes clinical knowledge executable.

We’re working with an early-stage, deeply technical clinical AI company building a fundamentally different approach to how AI can be used in healthcare.

Their core technology transforms clinical standards including prose, tables, algorithms, exceptions, and cross-references into executable, evidence-bound decision artifacts that are deterministic, verifiable, and traceable to their source evidence.

The technology is already live in clinical production, meaning this is an opportunity to work on frontier AI research with genuine real-world impact, not another research project sitting in a demo environment.

WHY THIS IS A GREAT OPPORTUNITY

Frontier AI + real-world clinical impact: Research how far modern models can be pushed while ensuring outputs meet the precision required in healthcare.
Research + production: Design experiments, build benchmarks, investigate model failures, and ship successful approaches into a production pipeline.
High ownership: Join a very small, highly technical team where your work can materially influence the product and research direction.
Work directly with founders and clinicians: Collaborate closely with the people building and validating the technology.
Remote-first: Work remotely from the US or Canada, with periodic in-person offsites.
Meaningful equity: Competitive equity package alongside a strong base salary.

WHAT YOU'LL DO

Build the pipeline that converts clinical standards into executable modules.
Develop and iterate on representations for clinical guideline logic.
Experiment with open/closed models, fine-tuning, and constrained decoding.
Design benchmarks, datasets, metrics, and adversarial cases for clinical fidelity.
Turn model failures into falsifiable hypotheses and experiments.
Develop deterministic validators using static analysis, constraint solving, and formal methods.
Work directly with clinicians and technical leadership to ensure outputs are complete, correct, and traceable.

WHAT WE'RE LOOKING FOR

You could be a strong fit if you have 3+ years of research or industry experience, or are a final-year PhD/postdoc, particularly in areas involving ML, AI, clinical AI, NLP, formal methods, or related fields.

The ideal candidate is someone who enjoys asking: “How do we know this actually works?”-and then designing the experiment, benchmark, or system needed to prove it.

Interested in building AI systems where correctness actually matters? Apply to learn more.