Researcher, Engineering, Systems Engineer

Overview

The company is building Sim Ops for Hardware
—an automated infrastructure layer that validates mechanical designs against physics constraints (thermal, fluid, structural) in real-time.

Today, simulation is a bottleneck because it is manual. Senior engineers spend hours cleaning geometry, defining mesh parameters, and waiting for solvers, only to find a basic error. We are automating this entire loop. We ingest raw, often imperfect manufacturing files (.stp), automatically repair and mesh them, and run deterministic physics solvers (OpenFOAM, proprietary kernels) to provide instant "Pass/Fail" triage to hardware teams.

We are the CI/CD pipeline for physical product development.

You will architect and implement the core physics engine that powers the company. Your primary technical challenge is robustness without human intervention
. You must build systems that can accept arbitrary, "dirty" geometry from a customer and return a converged simulation result without a user ever clicking "repair" or setting a mesh size.

• Automated Mesh Generation: Develop and refine algorithms to handle non-manifold, non-watertight, or complex assembly geometry. You will implement logic that automatically determines mesh density, boundary layers, and refinement zones based on feature detection (e.g., identifying heat sinks vs. chassis walls).
• Headless Solver Orchestration: Build the backend logic to drive OpenFOAM and other solvers programmatically. This includes automatic convergence monitoring, divergence recovery strategies, and dynamic timestep adjustment.
• Geometric Kernel Development: Work directly with B-Rep and mesh data structures (using tools like OCCT, Gmsh, or custom C++ implementations) to perform boolean operations, volume extraction, and surface repair at scale.
• Validation & Benchmarking: Create the case suites that verify our automated results against industry standards (Ansys/Fluent) to ensure physics fidelity.

We are looking for an engineer who understands the mathematical fundamentals of simulation, not just how to use a GUI. You should be able to look at a diverged residual plot and identify whether the issue is poor mesh quality, aggressive under-relaxation, or an ill-posed boundary condition.

• Background: MS or PhD in Engineering Physics, Applied Mathematics, Computer Science, or Mechanical Engineering.
• Core Tech: Expert proficiency in Python (for orchestration) and C++ (for core numerics/meshing).
• Simulation
  
  Experience:
  
  Deep experience with OpenFOAM
  , Ansys Fluent
  , or Star-CCM+. You understand the internal file structures (dictionaries, schemes) and have likely written custom scripts or user-defined functions (UDFs).
• Geometry: Familiarity with computational geometry concepts (NURBS, tessellation, isosurface extraction).
  
  Experience with kernels like Parasolid
  , ACIS
  , or Open CASCADE is a strong differentiator.

• Experience writing custom meshers or modifying open-source meshing libraries (Netgen, Gmsh).
• Background in High-Performance Computing (HPC), specifically optimizing solver performance on AWS/EC2 or Dockerized environments.
• Previous work on "verticalized" simulation tools (e.g., automated wind tunnel apps, internal tools for turbo machinery).