Job Description:

• Serve as the first ML hire on Platform Engineering and report to the Head of Platform Engineering
• Own the full lifecycle of extraction models: framing, data and annotation strategy, model selection, training/fine-tuning, evaluation, deployment, monitoring, and retraining
• Define “accurate enough” with clinical and customer-facing stakeholders
• Build the evaluation harness as an initial deliverable
• Partner with Clinical Data Managers on curation design
• Own the technical half of QA/QC, including extractability decisions, model specifications, tooling, sampling, and error analysis for human-in-the-loop review
• Build clinical NLP pipelines against messy real-world data
• Work with backend engineers to productionize ML systems
• Develop and rigorously evaluate LLM solutions using versioned prompts, cost and latency tracking, and failure-mode analysis
• Make extraction quality understandable to non-ML colleagues
• Handle de-identification, PHI, audit trails, and access controls as part of the modeling problem
• Help shape the roadmap and potentially lead the future team

Requirements:

• Healthcare or life sciences experience with real clinical data — clinical notes, EHR data, claims, registries, or similar; this is non-negotiable
• 6+ years in applied ML
• Personally taken models from problem statement to production and maintained them
• Depth in applied ML on text, including information extraction, NER, classification, sequence labelling, weak supervision, annotation guidelines, and inter-annotator agreement
• Practical experience with LLM-based approaches, including prompt development, structured output, retrieval, fine-tuning where warranted, evals, and observability
• Strong engineering fundamentals in Python
• Production engineering experience beyond notebooks
• Ability to define problems with stakeholders, write clearly, and collaborate across the ML boundary
• Self-directed problem-solving and comfort working without a playbook
• Nice to have: SNOMED CT, ICD-10, LOINC, RxNorm, CPT, FHIR
• Nice to have: HIPAA, SOC 2, de-identification methodology, IRB, or regulatory-grade data experience
• Nice to have: early or first ML hire experience
• Nice to have: human-in-the-loop annotation and QC operations at scale
• Nice to have: OCR and document-understanding experience
• Nice to have: mentoring or leading ML engineers
• Applicants must be located in the United States
• Novellia is not able to sponsor visas at this time

Benefits:

• Equity in Novellia
• Medical, dental, and vision coverage
• 401(k)
• Flexible time off
• Wellness stipend
• Up to 12 weeks of parental leave