Lead ML Platform Engineer — Scale & Self-Serve AI

Datavant; Charleston (WV); Vor Ort; USD 224.000 - 280.000; Vollzeit

Zusammenfassung

Datavant seeks a Staff ML Platform Engineer to lead the Data & ML Platform team, driving the paved road from model to production. You’ll own training, serving, observability, and architecture for LLM endpoints across Databricks, AWS, and self-hosted deployments, enabling self-serve workflows for Data Science teams.

Strong background in ML infrastructure, Spark, AWS, and CI/CD is required, along with a pragmatic approach to safety, cost, and performance when handling PHI and regulated data.

Qualifikationen

• 10+ years of software engineering experience with 3+ years designing ML platforms.
• Hands-on with Databricks and/or SageMaker and MLflow or equivalent.
• Fluency in Java and Python with Apache Spark expertise.
• Deep AWS knowledge: networking, IAM, GPU compute, storage, messaging.
• Terraform, containers, Kubernetes, and GitHub CI/CD for ML workloads.
• Experience serving LLMs in production with cost management and safety controls.

Aufgaben

• Set technical direction across ML training, serving, and observability; escalate rare infrastructure incidents.
• Evolve the paved-road framework for production workflows and Data Science self-service.
• Lead architecture for LLM-endpoint serving across managed and self-hosted deployments.
• Own standards for MLflow, model registry, training image supply chain, and observability.
• Partner with Data & ML Platform teams to present a cohesive platform to stakeholders.
• Mentor senior engineers and contribute code and IaC alongside teammates.

Kenntnisse

Software engineering

ML platform

Databricks

SageMaker

Python

Java

Apache Spark

AWS

Terraform

Kubernetes

CI/CD

LLM serving

Tools

MLflow

GitHub CI/CD

Databricks Asset Bundles

Unity Catalog

Iceberg/Delta

Jobbeschreibung

Datavant seeks a Staff ML Platform Engineer to lead the Data & ML Platform team, driving the paved road from model to production. You’ll own training, serving, observability, and architecture for LLM endpoints across Databricks, AWS, and self-hosted deployments, enabling self-serve workflows for Data Science teams.

Strong background in ML infrastructure, Spark, AWS, and CI/CD is required, along with a pragmatic approach to safety, cost, and performance when handling PHI and regulated data.