Applied Scientist, AI for Engineering Simulations

Mistral.ai; Paris; Sur place; EUR 90 000 - 140 000; Plein temps

Résumé du poste

Mistral AI seeks Applied Scientists to push AI-Physics integration across simulation domains. You will design and run large-scale simulations, train models, and build automation tools, with cross-functional collaboration to deliver production-grade AI solutions for engineering teams.

Ideal candidates hold a PhD or Master’s in AI or engineering science, strong Python and Linux/HPC skills, and experience with PyTorch or JAX. Fluency in English required; curiosity and teamwork valued.

Qualifications

• PhD or Master's in AI or engineering science is required.
• Fluent English and strong communication are essential.
• Proficiency in Python and Linux/HPC environments is expected.
• Experience with industry-grade projects or research studies is a plus.

Responsabilités

• Design and run large-scale simulation campaigns using domain-specific solvers.
• Train AI models on physics data with rigorous evaluation.
• Build tools for automated dataset creation and simulation pipelines.
• Develop agents and RAG integrating LLMs with engineering workflows.
• Collaborate with science/research teams on training runs and diagnose data issues.
• Manage research projects and client communications with engineering teams.

Connaissances

Fluent English

Python

Strong communication

Formation

PhD or Master’s in AI or engineering science (Mechanical/Electrical/CFD/Structural/semiconductor)

Outils

PyTorch

JAX

Description du poste

Mistral AI seeks Applied Scientists to push AI-Physics integration across simulation domains. You will design and run large-scale simulations, train models, and build automation tools, with cross-functional collaboration to deliver production-grade AI solutions for engineering teams.

Ideal candidates hold a PhD or Master’s in AI or engineering science, strong Python and Linux/HPC skills, and experience with PyTorch or JAX. Fluency in English required; curiosity and teamwork valued.