Staff Machine Learning Engineer

Role Overview

Fullscript is hiring a Staff Machine Learning Engineer to join its AI team. This is a senior individual contributor role where you will define technical direction and build production LLM-powered experiences for clinical workflows.

Responsibilities

• Design, develop, and deploy production, multi-turn LLM features (e.g., summarization tools and clinician-facing conversational agents)
• Own backend services in Python that integrate LLM agents with Fullscript’s platform for reliable production use
• Define technical direction for prompting, grounding, safety, and orchestration strategies across clinical AI workflows
• Establish and improve evaluation for LLM outputs, including accuracy, hallucinations, edge cases, and overall feature quality
• Shape engineering patterns for model workflows, including testing, CI/CD, observability, and version control
• Partner with engineering, product, analytics, and medical stakeholders to identify high-value AI opportunities and turn them into scalable product capabilities
• Provide technical leadership by clarifying ambiguous problem spaces, guiding tradeoffs, and raising the team’s quality bar
• Stay current with LLM research and emerging AI technologies and assess where they can be applied at Fullscript

Requirements

• 6+ years building and deploying machine learning applications in production, including meaningful experience with LLM-powered agents and conversational or agent-based workflows
• Demonstrated ability to own complex problems end to end and influence implementation beyond code
• Experience designing systems that handle open-ended questions, support follow-up interactions, and operate reliably in production
• Strong familiarity with LLM tooling/frameworks such as LangChain, LangGraph, or comparable orchestration and RAG frameworks
• Knowledge of evaluation and monitoring for LLM outputs, conversational quality, and system reliability
• Familiarity with agent orchestration patterns (including mentions of MCP)

Nice-to-haves

• Experience with multi-step agent workflows and related tooling/patterns (as suggested by knowledge of MCP and orchestration approaches)