{
  "job_id": "oLM1xW7l5Q5rydz5AAAAAA==",
  "job_city": "Nantes",
  "job_state": null,
  "job_title": "AI R&D Engineer",
  "job_salary": null,
  "job_country": "FR",
  "job_benefits": null,
  "job_latitude": 47.218371,
  "job_location": "Nantes",
  "job_onet_soc": "15111100",
  "apply_options": [
    {
      "is_direct": false,
      "publisher": "Trabajo.org",
      "apply_link": "https://fr.trabajo.org/offre-2554-8cfd8132bb7d6788bdd54bccee4b36f4?utm_campaign=google_jobs_apply&utm_source=google_jobs_apply&utm_medium=organic"
    },
    {
      "is_direct": false,
      "publisher": "Sercanto",
      "apply_link": "https://www.sercanto.fr/detail/a/ai-r-d-engineer_nantes_188281250?utm_campaign=google_jobs_apply&utm_source=google_jobs_apply&utm_medium=organic"
    },
    {
      "is_direct": false,
      "publisher": "Jobilize",
      "apply_link": "https://www.jobilize.com/job/fr-nantes-ai-r-d-engineer-deepmath-hiring-now-job-immediately-opening?utm_campaign=google_jobs_apply&utm_source=google_jobs_apply&utm_medium=organic"
    },
    {
      "is_direct": null,
      "publisher": "Trabajo.org",
      "apply_link": "https://fr.trabajo.org/offre-2554-8cfd8132bb7d6788bdd54bccee4b36f4"
    }
  ],
  "employer_logo": null,
  "employer_name": "deepmath",
  "job_is_remote": false,
  "job_longitude": -1.553621,
  "job_posted_at": "il y a 11 jours",
  "job_publisher": "Trabajo.org",
  "job_apply_link": "https://fr.trabajo.org/offre-2554-8cfd8132bb7d6788bdd54bccee4b36f4?utm_campaign=google_jobs_apply&utm_source=google_jobs_apply&utm_medium=organic",
  "job_highlights": {},
  "job_max_salary": null,
  "job_min_salary": null,
  "job_description": "We are deepmath, a deep-tech startup headquartered in Nantes, France, with operations in Brazil and strategic partnerships worldwide. We are excited to announce a CDI position for an AI R&D Engineer, starting March/April 2026, focused on developing advanced machine learning techniques with an emphasis on transformers and graph neural networks. If you are driven by innovation and motivated to tackle challenging problems in deep learning, we invite you to apply and join our journey\n\n1 About Us\n\nWe combine cutting-edge techniques in mathematical modeling, physics-based simulation, and artificial intelligence to deliver accurate descriptions and predictions of physical phenomena. Although we support projects in wind, solar, and offshore energy, our main focus is developing next-generation engineering simulation tools, where advanced physics and AI converge to tackle demanding industrial challenges.\n\ndeepmath is building a diverse, healthy, and supportive work environment where creativity thrives. We believe that great ideas emerge when people feel respected, supported, and empowered. Our goal is to create the conditions for every team member to grow, express their strengths, and reach their full potential.\n\n2 The Project\n\nEngineering simulation has long faced a critical bottleneck: the generation of high-quality meshes, especially for Computational Fluid Dynamics (CFD). For decades, this process has remained largely manual, time-consuming, and dependent on expert knowledge, making it one of the main obstacles to the large-scale industrial adoption of advanced simulation tools.\n\nOur project, deepmesh, addresses this challenge by developing an intelligent AI-driven mesh generator. By combining graph neural networks and transformer architectures, deepmesh learns to automatically produce simulation-ready meshes for complex geometries and demanding physical conditions. This technology enables faster, more reliable, and more accessible simulations, allowing engineers to focus on innovation rather than preprocessing, and laying the foundation for next-generation engineering design and optimization workflows.\n\n3 Your Missions\n\nAs an AI R&D Engineer, you will contribute directly to the evolution of our intelligent mesh generation system. Working closely with our CTO Bruno and our tech-dev team, you will help design, implement, and optimize the machine learning models that power deepmesh. Your role spans model architecture, data processing, large-scale training, and experimental validation.\n\nYour missions will include:\n• Developing and improving AI models based on graph neural networks and transformer architectures;\n• Designing and maintaining efficient data-loading, preprocessing, and batching pipelines for large-scale datasets;\n• Running extensive training experiments, analyzing results, and iterating to improve model accuracy, robustness, and performance;\n• Contributing to model optimization, training scalability, inference efficiency, and memory management;\n• Collaborating with the team to integrate new model components into the broader deepmesh pipeline.\n\nBy joining deepmath at an early stage, you will be in a position to make a significant impact not only on the technical success of the project but also on the direction of our technology strategy and future developments.\n\n4 Your Profile\n\nHard Skills\n• Graduate degree (MSc or PhD) in Computer Science, Applied Mathematics, Engineering, or a related field, with work involving deep learning or scientific machine learning;\n• Solid experience with modern deep learning architectures, especially transformers and graph neural networks;\n• Strong programming skills in Python and PyTorch and proficiency with Linux environments. Experience with simulation tools is a plus;\n• Experience with model development pipelines: dataloading, preprocessing, supervised training, evaluation, and optimization;\n• Experience with large-scale training on GPUs or cloud/HPC environments.\n• Proactivity and the",
  "job_google_link": "https://www.google.com/search?q=jobs&gl=fr&hl=fr&udm=8#vhid=vt%3D20/docid%3DoLM1xW7l5Q5rydz5AAAAAA%3D%3D&vssid=jobs-detail-viewer",
  "employer_website": "https://deepmath.tech",
  "job_onet_job_zone": "5",
  "job_salary_period": null,
  "job_apply_is_direct": false,
  "job_employment_type": "À plein temps",
  "job_employment_types": [
    "FULLTIME"
  ],
  "job_posted_at_timestamp": 1773100800,
  "job_posted_at_datetime_utc": "2026-03-10T00:00:00.000Z"
}