Senior AI Researcher- Pre-training (f/m/d)

OUR MISSION Aleph Alpha is one of the few companies in Europe doing serious foundation model pre-training. Our customers — in finance, manufacturing, and public administration — need models that understand German, meet European regulatory requirements, and work reliably in high-stakes settings. We’re building that in Heidelberg. We are hiring a Senior AI Researcher to join our Pre-training team and to advance the architecture and training of our next generation of foundation models. If you are excited about designing inference-efficient architectures, optimising training recipes that scale reliably, and training models on a large scale cluster (thousands of NVIDIA Blackwell GPUs), we would love to hear from you. TEAM CULTURE We foster a culture built on ownership, autonomy, and empowerment. Teams and individual contributors are trusted to take responsibility for their work and drive meaningful impact. We maintain a flat organisational structure with efficient, supportive management that enables quick decision-making, open communication, and a strong sense of shared purpose. We collaborate closely on complex technical problems, working in pairs or using mob programming to resolve challenging issues. ABOUT THE ROLE As a Senior AI Researcher in Pre-training (f/m/d), you will own the critical technical levers that determine the success of our next-generation models: architecture, optimization, stability, and scaling. Working at the high-leverage intersection of research and engineering, you will translate mathematical reasoning and empirical observations into principled training decisions - from small-scale proxy experiments to multi-thousand-GPU runs. We are looking for an expert who can combine rigorous experimental design with high-quality production code, directly influencing model quality, run reliability, and the efficiency of the models we ship. YOUR RESPONSIBILITIES - Recipe & Architecture Optimization: Own core elements of the training recipe (optimizers, schedules, initialization) and design PyTorch-based architectural improvements to maximize convergence, stability, and training efficiency. - Scaling Strategy & Predictability: Develop hyperparameter scaling laws and scale-up methodologies, using small-scale proxy experiments to reliably predict multi-thousand-GPU behavior and de-risk major training decisions. - Stability, Diagnostics & Debugging: Investigate complex convergence issues (loss spikes, divergence) and resolve hard-to-reproduce distributed system failures like communication bottlenecks, race conditions, and synchronization errors. - System-Model Co-Design: Partner with Compute Performance, Data, Evaluation, and Post-Training teams to align the model lifecycle with hardware constraints, memory bandwidth, and communication topologies. CORE QUALIFICATIONS - You are proficient in Python and deeply familiar with PyTorch-based training workflows. - You have a strong track record in machine learning research and software engineering, demonstrated through shipped models, impactful open-source contributions, or published research. - You have a strong mathematical foundation and are comfortable reasoning formally about optimisation, scaling behaviour, and training dynamics. - You deeply understand transformer training dynamics, optimisation, and the behaviour of large distributed training jobs. - You can design rigorous experiments, reason clearly from noisy results, and translate empirical observations into robust training decisions. - Hands-on experience pre-training large models (e.g., 7B+ parameters) on substantial infrastructure (e.g., 100+ GPU clusters). - You apply strong software engineering practices, including writing maintainable, well-tested code and supporting reproducible experimentation workflows. - You are able to implement complex model architectures efficiently and reliably and to debug complex issues across model code, training dynamics, and distributed systems. - You collaborate effectively within a research and engineering team and communicate clearly about your work across Pre-training and the broader AAR/AA organization. - You are able to work in Germany and collaborate regularly on site in Heidelberg as part of the Pre-training team. PREFERRED QUALIFICATIONS (We encourage you to apply even if you don't check every box!) - Large-Scale Training: Hands-on experience training LLMs or multimodal models on large GPU clusters using distributed frameworks (e.g., Megatron-LM, DeepSpeed, torchtitan). - Predictive Scaling: Familiarity with scaling laws, hyperparameter transfer, or methods for predicting large-scale training behavior from smaller proxy runs. - Stability & Performance: Experience profiling distributed jobs and diagnosing training anomalies like loss spikes, numerical instability, or optimizer pathologies. - Advanced Architectures: Exposure to sparse training approaches (e.g., Mixture-of-Experts) and an understa