Senior ML Platform Engineer – Production-Grade Pipelines

Datavant; Juneau (AK); On-site; USD 224,000 - 280,000; Full time

Job summary

Datavant is hiring a Staff ML Platform Engineer to lead the ML Platform team. You will set the technical direction for training, serving, and observability, and own the paved-road framework so data science teams can deploy workflows without bespoke plumbing.

You will architect LLM-endpoint serving across managed and self-hosted deployments, manage MLflow tooling, and guide the team with hands-on IaC across AWS, Spark, and containerized environments.

Qualifications

• 10+ years of software engineering experience with enterprise-scale ML platforms.
• Proven track record setting standards and guiding cross-functional teams.
• Hands-on production experience with Databricks and SageMaker, MLflow or equivalent.
• Fluency in Java and Python, and deep Spark expertise for large-scale data.
• Strong AWS know-how and Terraform, containers, Kubernetes for ML workloads.
• Experience deploying LLMs in production with safe handling of data.

Responsibilities

• Set technical direction for ML training, serving, and observability, and resolve complex infra issues.
• Evolve paved-road framework for scalable production workflows across DS teams.
• Lead architecture for LLM-endpoint serving across providers and self-hosted deployments.
• Own tools and standards for MLflow, model registry, and training image supply chain.
• Collaborate with Data & ML Platform teammates to present a cohesive platform.
• Influence vendor/platform choices for ML tooling and observability.
• Mentor engineers and contribute hands-on IaC and high-leverage code.

Skills

ML Platform

Databricks

SageMaker

Java

Python

Apache Spark

AWS

Terraform

Kubernetes

LLMs in production

Communication

Tools

Databricks Asset Bundles

Unity Catalog

Iceberg

Delta

Job description

Datavant is hiring a Staff ML Platform Engineer to lead the ML Platform team. You will set the technical direction for training, serving, and observability, and own the paved-road framework so data science teams can deploy workflows without bespoke plumbing.

You will architect LLM-endpoint serving across managed and self-hosted deployments, manage MLflow tooling, and guide the team with hands-on IaC across AWS, Spark, and containerized environments.