Forward Deployed Machine Learning Engineer

About the role

Triomics is creating AI agents that read oncology electronic health records at scale. As a Forward Deployed Machine Learning Engineer you will design, build and launch extraction pipelines that turn unstructured clinical documents into structured data for pharmaceutical partners and cancer hospitals.

Key responsibilities

• Design end‑to‑end agentic pipelines that process large patient charts (clinical notes, pathology reports, imaging and genomic panels) and output JSON matching customer data dictionaries.
• Define evaluation datasets, run precision/recall analyses, identify failure modes and improve accuracy through model tuning, prompt engineering or rule‑based post‑processing.
• Read and interpret raw patient charts to understand oncologists’ documentation practices and resolve ambiguous data points.
• Collaborate with annotation teams to create gold‑standard datasets and with customers to clarify dictionary definitions.
• Work with internal engineering and infrastructure teams to deploy, scale and monitor pipelines in production.
• Deliver on tight customer timelines, managing sprint cycles and iterative improvement phases.

Required profile

• Mid‑level experience (approximately 2 years) in machine‑learning engineering.
• Ability to dive deep into clinical data and translate domain knowledge into technical solutions.
• Strong communication skills for cross‑functional coordination with customers and internal teams.

Required skills

• Machine learning
• Prompt engineering
• Fine‑tuning of language models
• Rule‑based post‑processing
• Experience with JSON data structures