About The Position

Engineering simulation is undergoing a structural shift. What used to be a slow, solver-driven process is becoming a fast, exploration-driven workflow, enabled by surrogate models, physics-informed AI, and GPU-accelerated methods. This changes not just how simulations run, but how engineers design—moving from evaluating a few options to exploring entire design spaces. We are looking for a hands-on Applied AI Engineer to experiment with these emerging technologies. Your role is to try new tools, test them, benchmark them, integrate them into prototypes, determine what actually works in practice, and give an informed decision on its viability in Bentley’s products. The goal is to develop a clear understanding of where these approaches can realistically transform engineering workflows—and where they cannot. This is an experimental role. You will spend most of your time building and testing prototypes, not writing reports or maintaining production systems.

Requirements

  • Strong programming ability (Python required)
  • Experience with PyTorch, JAX, or similar ML frameworks
  • Solid understanding of numerical simulation: FEM, CFD, or multiphysics modeling
  • Familiarity with: Deep learning, surrogate modeling, reduced-order models, or optimization workflows
  • Good communication skills and collaborative mindset

Nice To Haves

  • Experience building experimental or research prototypes
  • Understanding of design space exploration (DOE, parameter studies, combinatorial optimization)
  • Exposure to physics-informed ML or scientific machine learning
  • Experience working with GPU or high-performance computing environments
  • Experience working with machine learning for physics simulations
  • Hands-on experience developing Multiphysics simulation code

Responsibilities

  • Keep track of new technologies in the field
  • Evaluate emerging simulation and AI tools through hands-on prototyping
  • Build and benchmark surrogate models against traditional FEM/CFD workflows
  • Test robustness, limitations, and failure modes of new approaches
  • Prototype hybrid pipelines (e.g., surrogate + high-fidelity validation)
  • Explore how these approaches can integrate into real engineering workflows
  • Report to stakeholders with clear recommendations
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service