Postdoctoral Researcher: Scalable AI for Power Grids

UT-Battelle; Oak Ridge (TN); On-site; USD 65,000 - 80,000; Full time

Benefits offered by this job

Medical and retirement plans

Flexible work hours

On-site amenities

Job summary

ORNL is seeking a Postdoctoral Research Associate to develop and scale AI/DL models for power grid systems, including surrogate models and OPF workflows. You will build HPC-ready training and inference pipelines on leadership-class systems and contribute to data-driven analysis of large-scale simulations.

The role requires a PhD in a relevant field, strong PyTorch experience, and a track record of publishing.

Qualifications

• PhD in computer science or an AI-related field completed within the last 5 years.
• Expertise in scalable deep learning, including distributed training and large-scale inference with PyTorch.
• Experience with HPC systems and multi-node workflows on CPU/GPU clusters.
• Data management for AI/ML workflows on large scientific datasets.
• Experience writing SLURM and PBS job submission scripts for HPC clusters.
• Proficiency in Linux, bash, Git, Python and reproducible software environments.
• Strong software engineering practices and publication record.

Responsibilities

• Design, implement, and deploy scalable AI/DL models for power grid systems.
• Develop HPC-ready workflows for distributed training and large-scale inference.
• Write Linux/bash and SLURM/PBS scripts for multi-node GPU/CPU workflows.
• Publish in journals/conferences and present technical work at events.
• Collaborate with a multidisciplinary team on AI/DL research supporting the Laboratory’s missions.

Skills

Scalable DL

Distributed training

HPC workflows

Python & Bash

SLURM & PBS scripting

PyTorch

Education

PhD in CS/AI field

Tools

Git

Python

Linux

SLURM

Job description

ORNL is seeking a Postdoctoral Research Associate to develop and scale AI/DL models for power grid systems, including surrogate models and OPF workflows. You will build HPC-ready training and inference pipelines on leadership-class systems and contribute to data-driven analysis of large-scale simulations.

The role requires a PhD in a relevant field, strong PyTorch experience, and a track record of publishing.