AI Platform Engineer: Scale GPU AI for Researchers

The Johns Hopkins University Applied Physics Laboratory; Laurel (MD); On-site; USD 85,000 - 165,000; Full time

Job summary

The Johns Hopkins University Applied Physics Laboratory (APL) seeks an AI Solutions Engineer to bridge researchers and a scalable AI computing platform. You will enable AI workloads from experimentation to large-scale GPU execution, working with Jupyter, containers, Kubernetes, and AI frameworks.

You will collaborate with researchers to move AI workloads into production, configure platform services, and optimize GPU utilization across distributed environments.

Qualifications

• Bachelor of Science degree or equivalent years of related professional work experience.
• At least three (3) years experience using container-based orchestration (e.g. Kubernetes) and run-time environments (e.g. Docker).
• Hands-on experience with Jupyter environments and GPU-accelerated AI frameworks such as PyTorch, TensorFlow, Hugging Face, or similar technologies.
• Working knowledge of Kubernetes and GPU computing concepts sufficient to configure, integrate, and troubleshoot AI applications running in shared computing environments.
• Experience supporting researchers or developers with model training, fine-tuning, inference, workload optimization, and troubleshooting.
• Proficiency with Python and shell scripting for automation, troubleshooting, and platform integration.
• Working knowledge of machine learning concepts and model development workflows.
• Strong problem-solving, communication, collaboration, prioritization, and continuous-learning skills.
• Eligibility to obtain Interim Secret level security clearance and ultimately Secret level clearance; U.S. citizenship required.

Responsibilities

• Own the technical configuration and evolution of the AI platform by evaluating new capabilities and establishing best practices.
• Partner with Linux and Kubernetes admins on deployments, upgrades, and troubleshooting across AI platform and computing environment.
• Provide hands-on technical assistance to researchers and data scientists using GPU-accelerated AI/ML platforms for model development, training, fine-tuning, and inference.
• Troubleshoot AI workloads across the stack including Python environments, Jupyter, containers, Kubernetes, GPU scheduling, storage, networking, and AI frameworks.
• Help researchers optimize GPU utilization and scale workloads from single-GPU to multi-GPU and multi-node distributed execution.
• Develop and maintain reusable environments, workflows, automation, documentation, and operational guidance to improve platform reliability and researcher productivity.
• Occasional weekend and after-hours work may be required to meet business needs.

Skills

Kubernetes

Docker

Python

Jupyter

GPU scheduling

PyTorch

TensorFlow

Shell scripting

ML concepts

Education

Bachelor's degree

Tools

NVIDIA CUDA

NVIDIA GPU Operator

NCCL

Kubeflow

MLflow

Weights & Biases

Job description

The Johns Hopkins University Applied Physics Laboratory (APL) seeks an AI Solutions Engineer to bridge researchers and a scalable AI computing platform. You will enable AI workloads from experimentation to large-scale GPU execution, working with Jupyter, containers, Kubernetes, and AI frameworks.

You will collaborate with researchers to move AI workloads into production, configure platform services, and optimize GPU utilization across distributed environments.