Member of Technical Staff, Compiler

About the role

A research and engineering role turning clinical standards into executable, verifiable decision logic, with the work running in live public health care rather than in demos.

What you will do

• Build a pipeline that parses clinical standards with document and layout models, then transforms and validates them into executable modules
• Define a target syntax for guideline logic covering exclusions, precedence and recommendations
• Run experiments comparing open and closed models, fine-tuning and constrained decoding
• Co-design benchmarks with datasets, metrics and error analysis for clinical fidelity
• Create adversarial cases testing source fidelity, logical completeness and recommendation correctness

What they look for

• Experience in a frontier AI lab
• Three or more years of research or industry work, or a final-year PhD or postdoc
• A track record of shipping ML or NLP systems to production
• A BS or higher in computer science, ML, math or a related technical field
• Production-level Python, PyTorch and Transformers experience
• Clear written communication about failed experiments and uncertainty

Nice to have

• Compilers, program synthesis, formal methods or static analysis
• Experience in healthcare, safety-critical or regulated domains
• A PhD in computer science, ML, NLP, formal methods or statistics
• Knowledge of IR design, constraint solving, SAT or SMT, or DSLs

Pay

USD 150,000 to 300,000 per year