AI Scientist, Agentic CAD (AI4Engineering)
Core
Teaching Mistral's LLMs to generate, analyze, and modify CAD designs (text-to-CAD, image-to-CAD, geometric analysis, QA, modification) and building agent architectures for multi-step CAD workflows.
Role type
Senior IC AI Scientist (Agentic CAD)
Builds
Pre- and post-training data pipelines, verifiers, evaluation benchmarks, and agent scaffolding for CAD tasks.
Domain
AI/ML applied to Computer-Aided Design (CAD) and Engineering
Deliverable
production ML models
Required skills
LLM development (SFT, RL, fine-tuning, GRPO), hands-on CAD experience (SolidWorks, CATIA, Siemens NX, FreeCAD, CADQuery, OpenCASCADE), Python, Linux/HPC, data pipeline scaling, agent architecture design, evaluation benchmark creation
Preferred skills
Agentic systems (tool use, multi-step planning, orchestration), reward modeling (RLHF/RLAIF/RLVR), first-author publications in agentic CAD/CAD-ML, open-source contributions, publications in engineering/ML venues (NeurIPS, ICLR, SIGGRAPH, etc.)
Technologies
Python, Linux, HPC, SolidWorks, CATIA, Siemens NX, FreeCAD, CADQuery, OpenCASCADE, build123d
Responsibilities
Design pretraining/SFT/RL data for CAD tasks; Define verifiers and evaluation criteria; Scale data pipelines for agentic CAD training; Design and improve agent architectures and harnesses; Build and maintain CAD evaluation benchmarks and diagnostic tooling; Collaborate with CAD domain experts to identify high-value workflows
Seniority
Senior, hands-on IC
