Senior Principal Machine Learning Engineer

Senior Principal Machine Learning Engineer

Job Location s: US-Remote

Job :

We are looking for a Senior Principal Machine Learning Engineer to lead the design and delivery of end-to-end ML/AI systems that turn vast volumes of claims, clinical, and member data into measurable performance and reduced waste. You will define technical strategy, drive cross‑functional alignment, and own systems that directly shape payment accuracy, risk adjustment, and quality outcomes for the payers we serve.

This role sits at the intersection of applied research and production engineering, translating ambiguous, high‑stakes problems into scalable, auditable ML solutions.

The ideal candidate has operated at large scope across multiple teams and product surfaces – not just shipped models, but defined the problem, built the evaluation infrastructure, created the data flywheel, and drove measurable business outcomes. They think in systems, write crisp design docs, bring intellectual honesty to experimentation, and treat auditability and precision as first‑class requirements rather than afterthoughts. They raise the level of the engineers around them.

• Define system architecture for AI/LLM‑powered products end to end over claims, medical records, and clinical documentation.
• Build and own evaluation frameworks (LLM‑as‑a‑Judge, offline metrics, online experiments) aligned to accuracy, auditability, and clinical and regulatory risk.
• Drive the data flywheel: convert expert clinician and auditor review decisions into high‑quality labeled data, and close the loop with fine‑tuning of models to lift detection precision.
• Explore building patient‑level digital twins from clinical charts for unified processing layer and data presentation across payment, risk and quality.
• Lead ranking and prioritization systems that surface the highest‑value claims, audits, and care gaps for human review, improving both reviewer efficiency and financial impact.
• Establish reusable platform patterns – shared context stores, evaluation harnesses, feature pipelines – that compound value across product surfaces and lines of business.
• Partner across engineering, product, clinical, and analytics teams to align on success criteria, roadmap priorities, and production rollout.
• Mentor senior engineers and elevate organization‑wide standards in ML craftsmanship, experimentation rigor, and system design.
• Set company‑wide standards.
• Act as a thought leader beyond Cotiviti to elevate the reputation and visibility of Cotiviti in the industry.
• Influence the enterprise AI/ML strategy at an executive level.
• Complete all responsibilities as outlined in the annual performance review and/or goal setting.
• Complete all special projects and other duties as assigned.
• Must be able to perform duties with or without reasonable accommodation.

This job description is intended to describe the general nature and level of work being performed and is not to be construed as an exhaustive list of responsibilities, duties and skills required. This job description does not constitute an employment agreement and is subject to change as the needs of Cotiviti and requirements of the job change.

• PhD in a quantitative discipline such as Computer Science/Engineering, Statistics, Operations Research covering Advanced Statistics, Machine learning and AI.
• 12+ years of industry experience building production ML systems at scale.
• Deep expertise in two or more of: LLM evaluation, retrieval‑augmented generation (RAG), ranking, or large‑scale classification.
• Proven track record leading end‑to‑end ML projects, from problem framing through production impact.
• Strong experimentation discipline: A/B testing, causal inference, metric design, and opportunity mining.
• Proficiency in Python (PyTorch), SQL at scale (Presto / Trino / Spark), and distributed pipeline tooling (Airflow).
• Demonstrated ability to drive cross‑functional alignment across engineering, product, and analytics.

• Experience building LLM‑as‑a‑Judge evaluation pipelines aligned to quality, risk, and accuracy criteria.
• Hands‑on supervised fine‑tuning of embedding or reranking models with measurable…