Senior MLOps Engineer — Scale LLM Infra, Equity, Impact

Nace.AI; Palo Alto (CA); Hybrid; USD 180.000 - 240.000; Vollzeit

Benefits dieser Stelle

Equity

Premium benefits

Zusammenfassung

Nace AI is building production-grade ML infrastructure that turns research into reliable models. As a Senior MLOps Engineer, you will own pipelines, training orchestration, model registries, and CI/CD for models across multi-GPU clusters.

You will enable low-latency inference with vLLM, ensure observability, security, and auditability, and set best practices for an early infrastructure team in a fast-paced, enterprise-ready environment.

Qualifikationen

• 5+ years of experience in MLOps or ML infrastructure with production ownership.

Aufgaben

• Design, build, and operate end-to-end ML infrastructure: training orchestration, experiment tracking, model registries, CI/CD for models, and automated evaluation pipelines.
• Own LLM/SLM serving infrastructure — scale low-latency, high-throughput inference using frameworks like vLLM.
• Build and manage multi-GPU training and inference clusters across cloud and on‑prem environments.
• Implement observability for models in production: latency, throughput, drift, regression, and quality monitoring with alerting.
• Apply inference-time optimizations — quantization, distillation, KV-cache management, deployment tuning — with ML/Research teams.
• Harden stack for enterprise deployment: reproducibility, versioning, access controls, audit-ready traceability.
• Set MLOps best practices and tooling standards as an early member of the infrastructure team.

Kenntnisse

MLOps

Python

Kubernetes

Docker

Terraform

Airflow

Kubeflow

Ray

MLflow

Weights & Biases

Ausbildung

BS in CS or related field

MS in CS or related field (preferred)

Tools

vLLM

TRT

SGLang

TGI

Jobbeschreibung

Nace AI is building production-grade ML infrastructure that turns research into reliable models. As a Senior MLOps Engineer, you will own pipelines, training orchestration, model registries, and CI/CD for models across multi-GPU clusters.

You will enable low-latency inference with vLLM, ensure observability, security, and auditability, and set best practices for an early infrastructure team in a fast-paced, enterprise-ready environment.