(Senior) AI Scientist (m/f/d)

eigenblue - We're building the future of engineering. A future where engineers build intelligent systems that solve complex problems, and work alongside those systems to bring their solutions into the real world. We solve complex, multidisciplinary engineering problems by combining deep human expertise with AI systems we build ourselves. Our proprietary AI models generate system architectures across software, electronics, mechanics, and applied physics - and our engineers stay in the loop, not as supervisors but as the decisive factor to bring those architectures into the real world.

As (Senior) AI Scientist (m/f/d) you'll help push the reasoning core of our proprietary system, our multidisciplinary engineering world model, beyond what today's LLM-based approaches can do.

Your Role

• Senior AI Scientist (m/f/d) working directly on engineering advanced AI systems and world model's reasoning algorithms
• Develop simulation environments and reinforcement learning, planning and control methods, for robotics and mechatronics systems
• Researching and developing neural generative reasoning over formal, ontology-constrained representations (transformer/diffusion-based architecture generation)
• Designing constraint-aware combinatorial and graph search methods that combine neural proposal operators with deterministic solvers and physics-based simulations
• Working on causal modeling, physical plausibility, and cross-domain constraint propagation within a typed engineering ontology
• Generating and validating synthetic, physically-grounded training data for rare, costly, or safety-critical engineering scenarios
• Publishing internally and, where relevant, externally, contributing to eigenblue's scientific credibility in the field

Your Profile

• PhD (or equivalent research experience) in Machine Learning, Computer Science, Physics, Applied Mathematics, or a related computational field
• Strong publication record or demonstrable research depth in advanced AI, generative modeling, reinforcement learning, world models, or graph-based reasoning
• Hands-on experience with transformer and/or diffusion architectures beyond natural language: structured, constrained, or multimodal generation
• Comfort working with formal representations: ontologies, typed graphs, constraint satisfaction, or symbolic-neural hybrid systems
• Solid software engineering practice (Python, PyTorch/JAX): you can take research from notebook to reproducible pipeline
• Excellent written and verbal communication skills in English, with the ability to collaborate effectively across multidisciplinary teams

eigenblue is colorful - We embrace and value equal opportunities, inclusion, and diversity. Therefore, we welcome every application with pleasure!

Does that make you curious? Julia is your contact person for further questions and looks forward to hearing from you!