Onsite SciML Postdoc: Data‑Driven Simulation ML

Sandia National Laboratories; California (MO); On-site; USD 75,000 - 95,000; Full time

Job summary

Sandia National Laboratories seeks a Postdoctoral Appointee to advance trustworthy scientific machine learning and data‑driven analysis for computational science applications. You will develop ML surrogates and reduced‑order models within physics‑based simulations and perform verification and uncertainty quantification studies.

You will collaborate with Sandia PIs and academic partners to benchmark methods against high‑fidelity models and experiments, with onsite work in Livermore, CA.

Qualifications

• PhD in applied mathematics, computational science, engineering, statistics, computer science, or related field.
• Experience developing or evaluating ML methods for simulations, images, signals or high‑dimensional data.
• Ability to obtain and maintain a DOE Q-level security clearance.

Responsibilities

• Develop and evaluate ML surrogates, reduced‑order models, and data‑driven components within physics‑based simulations.
• Design verification, validation, uncertainty quantification, robustness testing, explainability and interpretability studies.
• Apply feature extraction and pattern recognition methods to images, sensor data, and high‑dimensional outputs.
• Collaborate with PIs, national labs and academic partners to benchmark data‑driven methods against high‑fidelity models.
• Work onsite in Livermore, CA.

Skills

Machine learning

Scientific ML

Uncertainty quantification

Statistical learning

Python

ML libraries

Education

PhD (within 5 years)

Job description

Sandia National Laboratories seeks a Postdoctoral Appointee to advance trustworthy scientific machine learning and data‑driven analysis for computational science applications. You will develop ML surrogates and reduced‑order models within physics‑based simulations and perform verification and uncertainty quantification studies.

You will collaborate with Sandia PIs and academic partners to benchmark methods against high‑fidelity models and experiments, with onsite work in Livermore, CA.