Research Scientist I/ii, Multiscale & Multiphysics Simulations

Lila Sciences Lila Sciences · AI Frontier · Alewife, Cambridge, MA · Physical Sciences AI

Research Scientist role focused on building AI-driven multiphysics and multiscale simulation capabilities for scientific discovery. The role involves developing high-fidelity digital representations of physical systems and integrating them into autonomous discovery and experimental pipelines, with a focus on agent-driven simulation workflows and closed-loop systems.

What you'd actually do

  1. Develop and deploy robust multiphysics models across coupled domains (e.g., thermal, fluid, structural, electromagnetic, chemical), using methods such as coarse-grained, mesoscale, FEM, and CFD techniques.
  2. Build integrated multiscale frameworks that connect atomistic, mesoscale, and continuum representations to model materials and devices.
  3. Design and implement programmatic, agent-driven simulation workflows that can autonomously configure, execute, and refine simulations within closed-loop discovery workflows.
  4. Create scalable, GPU-accelerated simulation pipelines, data infrastructure, and interoperable APIs that connect commercial tools (e.g., COMSOL, ANSYS) and custom solvers deploying on cloud-based, high-throughput computing environments
  5. Collaborate with AI, software, and automation teams to orchestrate and deploy closed-loop discovery workflows, integrating computational predictions with robotic and cloud-based laboratory platforms to enable automated experiment–simulation feedback cycles and accelerated R&D.

Skills

Required

  • PhD in Mechanical Engineering, Chemical Engineering, Aerospace Engineering, Materials Science, or a related field.
  • Extensive experience with multiphysics simulation methods and numerical algorithms, including FEM, CFD, TCAD/process simulation, mesoscale modelling, or related techniques.
  • Strong foundation in coupled physical phenomena, including heat transfer, fluid dynamics, structural mechanics, mass transport, diffusion, electromagnetism, and reaction kinetics.
  • Experience applying simulation to real-world systems in industrial settings such as semiconductors, chemical processing, aerospace, or materials manufacturing.
  • Solid programming skills in Python and building simulation workflows, automation pipelines, or custom numerical models.

Nice to have

  • Experience bridging atomistic simulations with one or more additional simulation domains including coarse-grained, finite-element and continuum models.
  • Familiarity with machine learning approaches applied to physical simulations (e.g., surrogate models, neural operators, physics-informed neural networks), along with experience leveraging GPU acceleration and programmatic optimization for scalable simulations
  • Experience integrating simulation frameworks into digital twin systems, real-time decision environments, or closed-loop control workflows.
  • Background applying simulation to complex materials and process domains such as thin-film deposition, micro/nano-fabrication, or reactive transport, with an understanding of processing–structure–property relationships.

What the JD emphasized

  • agent-driven simulation workflows
  • closed-loop discovery workflows
  • autonomous discovery and experimental pipelines

Other signals

  • AI-driven scientific discovery
  • autonomous discovery and experimental pipelines
  • agent-driven systems
  • closed-loop workflows