MLOps Engineer - GPU & LLM Infrastructure

Trine Infotech; United States; Remote; USD 140,000 - 190,000; Full time

Job summary

Trine Infotech is seeking an experienced MLOps Engineer to design and build scalable infrastructure for AI/ML and LLM workloads from scratch. You will own end-to-end deployment pipelines, GPU-enabled infrastructure, and production model serving.

The role requires deep hands-on experience with Kubernetes, Docker, Terraform/Ansible/Pulumi, and cloud/Linux environments to optimize performance, cost, and reliability while collaborating with AI/ML engineers.

Qualifications

• 6+ years of MLOps or related infra experience.
• Hands-on with GPU provisioning for AI workloads.
• Production deployment of ML models and LLMs.
• Strong Kubernetes and container orchestration knowledge.
• Experience with IaC tools: Terraform, Ansible, Pulumi.

Responsibilities

• Design end-to-end MLOps infrastructure for training, deployment, monitoring, and inference.
• Build CI/CD pipelines for ML models and automate deployment, validation, and rollback.
• Provision, configure, and optimize GPU-enabled infrastructure for AI/ML workloads.
• Deploy and scale LLM inference and model-serving solutions in production.
• Manage Kubernetes, Docker, infrastructure-as-code, and distributed systems.
• Implement monitoring, logging, observability, and automated remediation for production workloads.
• Optimize GPU/CPU utilization, latency, performance, and costs.
• Collaborate with AI/ML engineers to build scalable AI platforms.

Skills

MLOps experience

GPU provisioning

Kubernetes

Docker

Infrastructure-as-code

Python or Go

ML CI/CD / model serving

Cloud infrastructure

Tools

Terraform

Ansible

Pulumi

Job description

Job Summary

We are looking for an experienced MLOps Engineer to build and manage scalable infrastructure for AI/ML and Large Language Model (LLM) workloads from scratch. The ideal candidate will have strong expertise in ML deployment pipelines, GPU infrastructure, Kubernetes, automation, and production model serving.

Key Responsibilities

• Design and build end-to-end MLOps infrastructure for model training, deployment, monitoring, and inference.
• Build CI/CD pipelines for ML models and automate model deployment, validation, and rollback.
• Provision, configure, and optimize GPU-enabled infrastructure for AI/ML and LLM workloads.
• Deploy and scale LLM inference and model-serving solutions in production environments.
• Manage Kubernetes, Docker, infrastructure-as-code, and distributed computing environments.
• Implement monitoring, logging, observability, and automated remediation for production AI workloads.
• Optimize GPU/CPU utilization, inference latency, system performance, and infrastructure costs.
• Collaborate with AI/ML engineers to build reliable, scalable, and high-performance AI platforms.

Required Skills

• 6+ years of experience in MLOps, ML Platform Engineering, SRE, DevOps, or Infrastructure Engineering.
• Strong hands-on experience with GPU provisioning, GPU-enabled infrastructure, and AI/ML workloads.
• Experience deploying and operating ML models and LLMs in production.
• Strong knowledge of Kubernetes, Docker, and container orchestration.
• Experience with Terraform, Ansible, Pulumi, or similar infrastructure-as-code tools.
• Proficiency in Python or Go for automation and infrastructure tooling.
• Experience with ML CI/CD, model serving, monitoring, Prometheus, Grafana, or OpenTelemetry.
• Strong understanding of cloud infrastructure, Linux, networking, and distributed systems.

Good to Have

• Experience with Ray, Slurm, vLLM, TensorRT-LLM, or Triton Inference Server.
• Exposure to GPU clusters, distributed training, NVIDIA GPU technologies, NVLink, or InfiniBand.
• Experience building AI infrastructure or MLOps platforms from scratch.

Preferred Candidate:

Someone who can independently design, build, deploy, and optimize production-grade AI/ML and LLM infrastructurenot just maintain existing pipelines.