MLOps Engineer, Health AI & Claims Automation

Alaffia Health; New York (NY); Hybride; USD 140 000 - 210 000; Plein temps

Avantages offerts par ce poste

Competitive Compensation

Medical, Dental, and Vision benefits

Flexible, paid vacation policy

Flat organizational structure — direct

Résumé du poste

Alaffia Health, a healthtech AI startup in New York, is hiring an MLOps Engineer to build the operational backbone behind our AI agents. You will grow a library of agentic systems across many health-plan customers, ensuring reliability, cost-efficiency, and measurable performance from model experiments to production deployments.

You will collaborate with AI engineers and clinicians, owning end-to-end systems, from model experiments in development to deployed, monitored agents in production.

Qualifications

• Five+ years in MLOps, ML platforms or ML infra engineering.
• Experience with LLM-based or agentic systems preferred.
• Strong Python and software engineering fundamentals.

Responsabilités

• Design, build, and own agent evaluation framework covering output quality, throughput, and cost.
• Integrate automated evals into CI/CD pipelines.
• Manage model access across cloud providers and onboard new models.
• Build experiment tracking, metrics, and observability for experiments in dev.
• Develop versioned pipelines for models, prompts, and agent configurations.
• Monitor production agents for quality, throughput, cost, and drift; build dashboards and alerts.

Connaissances

Python

MLOps

CI/CD

ML model development

Observability

Outils

AWS Bedrock

Azure OpenAI

SageMaker

MLflow

Weights & Biases

Description du poste

Alaffia Health, a healthtech AI startup in New York, is hiring an MLOps Engineer to build the operational backbone behind our AI agents. You will grow a library of agentic systems across many health-plan customers, ensuring reliability, cost-efficiency, and measurable performance from model experiments to production deployments.

You will collaborate with AI engineers and clinicians, owning end-to-end systems, from model experiments in development to deployed, monitored agents in production.