AI Scientist - Physics Simulation Models

About Mistral

Mistral delivers comprehensive AI solutions encompassing frontier AI models, developer tools, applications, and computing resources. We collaborate with enterprises addressing significant challenges across industries such as finance, manufacturing, defense, healthcare, and the public sector, developing tailored AI systems deployable on the clients' terms. Our dynamic and diverse teams across Europe, North America, Asia, and the Middle East foster a creative, team-oriented, and low-ego culture focused on innovation.

Role Overview

We seek AI Scientists with profound expertise in engineering sciences and machine learning to advance AI-driven simulation. In this position within the AI4Engineering Science group, you will innovate foundational physics models that outperform current capabilities and enable fine-tuning for varied downstream applications by clients and internal teams.

You will manage end-to-end research activities including curating detailed simulation datasets, architecting and training novel models, and performing rigorous validation aligned with engineering standards. Collaborating throughout the research organization, you will ensure models are versatile foundations for multiple products rather than isolated solutions.

Key Responsibilities

• Develop and train advanced physics simulation foundation models that exceed existing standards in precision, generality, and scalability.
• Orchestrate extensive simulation campaigns utilizing domain-specific solvers to construct high-resolution datasets vital for foundational models.
• Explore and experiment with model architectures and training methodologies—such as multi-fidelity training and pretraining objectives—aimed at enabling a single model to adapt efficiently across diverse engineering challenges.
• Conduct thorough evaluations of model performance, robustness, and coverage against industry validation benchmarks, identifying and analyzing failures linked to data or architectural constraints.
• Remain abreast of scientific advances, contributing to Mistral’s leadership in AI-powered engineering research.

Qualifications and Skills

• PhD or Master’s degree in Computer Science, Artificial Intelligence, Mechanical Engineering, Electrical Engineering, Computational Fluid Dynamics, Structural Mechanics, EDA, Semiconductor Engineering, or related disciplines.
• Extensive hands-on experience in machine learning with deep knowledge of model architectures, training processes, and validation techniques.
• Proven background in developing ML methods tailored for simulation or surrogate modeling tasks.
• Proficiency in writing clean Python code and experience in Linux/HPC environments.
• Excellent English communication skills, capable of conveying complex simulation and ML concepts to technical and non-technical stakeholders.
• Self-motivated with an ability to progress without detailed guidance.
• Collaborative mindset with humility and eagerness to learn at the intersection of simulation and ML.
• Demonstrated achievements in industry projects, academic research, or personal initiatives.

Preferred Experience

• Experience with simulation solvers such as OpenFOAM, LS-DYNA, ANSYS, COMSOL, Abaqus, Fluent, STAR-CCM+, PowerFlow, NekRS, Tau/CODA, JAX-Fluids, or equivalents; EDA tools like Cadence, Synopsys, or Siemens.
• Familiarity automating large-scale simulation workflows on HPC clusters.
• Contributions to major open-source or industrial software projects.
• Publications in reputable engineering or machine learning conferences and journals (AIAA, ASME, JFM, NeurIPS, ICLR).

Compensation & Benefits

We provide a comprehensive benefits package that supports employees' health, career growth, and work-life balance, varying by location. Benefits may include healthcare, parental leave, retirement plans, relocation aid, wellness initiatives, meal and transport allowances, among others. For specific details, consult the benefits information applicable to your location.

Privacy Note

Your personal information will be handled confidentially in accordance with our Applicant Privacy Policy.