PhD ML Intern - Surrogate Modeling for Engineering

Anite; Calabasas (CA); On-site; USD 51,000 - 56,000; Full time

Benefits offered by this job

Mentorship across ML/engineering

Access to engineering tools and real D

Job summary

Keysight invites PhD students to join a machine learning internship focusing on surrogate models for engineering simulation and measurement. You will design and train neural models, study efficient adaptation with LoRA, and integrate results into a C++/LibTorch runtime, collaborating with RF design, measurement, and runtime engineers.

Gain hands-on experience applying ML to RF/high-speed problems, with mentorship across ML, simulation, and engineering disciplines, and exposure to industrial

Qualifications

• PhD in ML/CS/Applied Math/EE or related field.
• Hands-on experience with multiple neural architectures (transformers, neural operators, graph nets, CNNs).
• Strong Python and PyTorch with training/debugging beyond tutorials.
• Strong C++ and LibTorch to implement and integrate models into a runtime.
• Solid grasp of transfer learning and parameter-efficient methods such as LoRA.
• Strong statistics and exploratory data analysis with sensitivity analysis.
• Rigorous baselines, ablations, experiment tracking, reproducible results.
• Willingness to learn underlying physics; mentorship provided.

Responsibilities

• Generate parametric engineering datasets via simulation/measurement with attention to sampling and noise.
• Design, train, and evaluate neural surrogate models for engineering response prediction.
• Investigate output representations across operating regions and frequencies for accuracy.
• Establish baselines and quantify accuracy vs. simulation/measurement budget.
• Evaluate LoRA and related methods for specializing models to new problems.
• Develop calibrated uncertainty estimates for trustworthy predictions.
• Verify physical constraints (passivity, causality, reciprocity) and report metrics.
• Analyze data quality, solver noise, and error concentration across frequency/design space.
• Prototype training/inference in C++/LibTorch within an engineering ML runtime.

Skills

Python

PyTorch

C++

LibTorch

Transfer learning

LoRA

Experimental design

Statistics

Education

PhD in Machine Learning, Computer Science, Applied Mathematics, Electrical or Electronic Engineering

Tools

LibTorch

CUDA

TensorRT

CMake

Git

Job description

Keysight invites PhD students to join a machine learning internship focusing on surrogate models for engineering simulation and measurement. You will design and train neural models, study efficient adaptation with LoRA, and integrate results into a C++/LibTorch runtime, collaborating with RF design, measurement, and runtime engineers.

Gain hands-on experience applying ML to RF/high-speed problems, with mentorship across ML, simulation, and engineering disciplines, and exposure to industrial