Senior ML Engineer

About this role

Own machine-learning projects from exploration through production and ongoing iteration. Build and evaluate agentic LLM systems for clinical care, improving quality and reliability across the production AI stack.

About the company

Sword Health is a digital musculoskeletal care company that helps employers and health plans deliver AI-powered physical therapy and related care programs.

What you'll do

• Own ML projects end-to-end, from exploration to production and iteration
• Build reliable agentic LLM workflows with tools, retrieval, and orchestration
• Develop evaluation datasets, offline and online harnesses, human-reviewed LLM judges, and regression tests
• Improve model quality using evidence-based prompting, retrieval, distillation, or fine-tuning
• Work across data preparation, model adaptation, serving, monitoring, and production feedback loops
• Partner with Product, Clinical, and Engineering stakeholders
• Review code, share knowledge, and mentor less experienced engineers

What we're looking for

• Experience shipping production ML systems that people depend on
• Production experience with LLM prompting, retrieval, tool calling, and agent-style workflows
• Experience building evaluation datasets and frameworks and distinguishing real improvements from noise
• Strong ML fundamentals and ability to reason about approach tradeoffs
• Ability to turn ambiguous problems into production systems
• Production-quality engineering skills, distributed-systems familiarity, and ML pipeline debugging experience
• Clear communication with technical and clinical stakeholders

Nice to have

• Fine-tuning or preference optimization experience, such as RLHF or DPO
• Healthcare AI or experience in other high-stakes domains
• Built agent frameworks or evaluation tooling from scratch
• Open-source contributions, technical writing, or other knowledge sharing