Postdoctoral Researcher - Optimization with Embedded Machine Learning Surrogates

ExxonMobil • Spring, TX

2d • Onsite

About The Position

ExxonMobil is seeking a highly motivated Postdoctoral Researcher specializing in the integration of mathematical optimization and machine learning through surrogate modeling. This role focuses on embedding ML-based surrogate models directly within optimization frameworks to enable efficient decision-making for large-scale, high-value business applications. A key challenge lies in balancing surrogate model fidelity with optimization tractability and developing scalable solution algorithms for resulting nonconvex and large-scale formulations. The ideal candidate is a recent Ph.D. graduate with strong expertise in operations research, mixed integer linear or nonlinear optimization, and machine learning, with interest in solving real-world industrial problems involving complex physical systems.

Requirements

• Ph.D. in Operations Research, Industrial Engineering, Applied Mathematics, or a closely related field.
• Strong background in mathematical optimization, including nonlinear and mixed-integer optimization.
• Demonstrated research experience in at least one of the following: Optimization with embedded machine learning models, Surrogate-based optimization, Nonconvex or bilinear optimization.
• Knowledge of machine learning models used for surrogate modeling (e.g., neural networks, regression models).
• Strong programming skills in Python.
• Experience with optimization solvers (e.g., Gurobi, CPLEX, IPOPT).
• Strong analytical, problem-solving, and communication skills.
• Ability to work in multidisciplinary teams with domain experts.

Nice To Haves

• Experience with tools such as GurobiML, OMLT, or similar ML-to-optimization frameworks.
• Experience with derivative-free optimization methods (e.g., NOMAD, Bayesian optimization).
• Knowledge of gradient-based nonlinear optimization methods (e.g., SLSQP).
• Experience working with large-scale industrial or engineering systems.
• Understanding of surrogate model training and validation trade-offs.
• Strong publication record.
• Experience developing reusable optimization frameworks or toolkits.

Responsibilities

• Develop optimization frameworks with embedded ML-based surrogate models for complex systems.
• Design and implement formulations that integrate neural networks and other surrogate models into optimization problems (e.g., MIP, MINLP, and nonconvex programs).
• Investigate trade-offs between surrogate model fidelity and optimization tractability.
• Develop specialized solution algorithms for challenging problem structures, including bilinear and nonconvex formulations.
• Explore hybrid solution approaches combining: Mathematical programming (e.g., MIP/MINLP), Gradient-based optimization (e.g., SLSQP), Derivative-free optimization (e.g., NOMAD).
• Leverage tools such as GurobiML, OMLT, and decomposition methods.
• Apply developed methods to high-impact business problems across upstream, downstream, and low-carbon solutions.
• Communicate results through technical reports, publications, and presentations.

Benefits

• Relocation benefits may be available to you based on ExxonMobil eligibility guidelines.

What This Job Offers

• Job Type: Full-time
• Career Level: Entry Level
• Education Level: Ph.D. or professional degree