This list contains only the countries for which job offers have been published in the selected language (e.g., in the French version, only job offers written in French are displayed, and in the English version, only those in English).
Domyn is a company specializing in the research and development of Responsible AI for regulated industries, including financial services, government, and heavy industry. It supports enterprises with proprietary, fully governable solutions based on a composable AI architecture — including LLMs, AI agents, and one of the world's largest supercomputers. At the core of Domyn's product offer is a chip-to-frontend architecture that allows organizations to control the entire AI stack — from hardware to application — ensuring isolation, security, and governance throughout the AI lifecycle. Its foundational LLMs, Domyn Large and Domyn Small, are designed for advanced reasoning and optimized to understand each business's specific language, logic, and context. Provided under an open-enterprise license, these models can be fully transferred and owned by clients. Once deployed, they enable customizable agents that operate on proprietary data to solve complex, domain-specific problems. All solutions are managed via a unified platform with native tools for access management, traceability, and security. Powering it all, Colosseum — a supercomputer in development using NVIDIA Grace Blackwell Superchips — will train next-gen models exceeding 1T parameters. Domyn partners with Microsoft, NVIDIA, and G42. Clients include Allianz, Intesa Sanpaolo, and Fincantieri.
Job Responsibility
Build the next generation of large language models, across the full life cycle: pretraining on trillions of tokens, state-of-the-art post-training, and releases spanning from a few billion to hundreds of billions of parameters
Collaborate directly with leading industry labs like NVIDIA
Work will ship into mission-critical deployments with some of the largest players in banking, defense, manufacturing, and the public sector
Design and optimize the infrastructure that trains and serves these models
Contribute to open-source projects shaping the European AI ecosystem
Requirements
PhD in Computer Science, Artificial Intelligence or a related field, or equivalent practical experience
At least 5 years of proven experience as an AI research engineer or more than 2 years of experience and a PhD
Expertise in modern machine learning frameworks such as PyTorch, TensorFlow, and JAX, with deep knowledge of distributed training (Pytorch Distributed, Ray, DeepSpeed)
Strong background in parallel computing and high-performance systems, including CUDA programming and compiler optimizations
Hands-on experience with ML model debugging and performance profiling tools (TensorBoard, Weights & Biases, NVIDIA Nsight)
Proficiency in Python and C++ or Rust, particularly for high-performance inference and AI accelerators
Solid understanding of mathematics behind deep learning (linear algebra, probability, optimization)
Experience deploying models in production, optimizing for latency, throughput, and memory efficiency
Fluent in English
Nice to have
Track record of publications at top-tier AI venues, reflecting a deep understanding of frontier research
Contributions to open-source AI projects, demonstrating systems-level thinking and impact in the research community
Knowledge of AI safety, alignment, or adversarial robustness techniques
Experience with low-precision training (FP8, FP4), quantization, pruning, or distillation at scale
What we offer
Learning Friday
Training budget for books, online courses or other training materials