Latest AI JobsBrowse Jobs
AlephAlpha logo

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

AlephAlphaHeidelbergResearch
Apply on company site
Salary
Not listed
Job type
Full-time
Work mode
Remote
Experience
Senior
Posted
9 Oct 2026
Status
Open · checked 7d ago

About this role

Based in Heidelberg, the Senior AI Researcher - Pre-training position at Aleph Alpha involves advancing the architecture and training recipes for next-generation foundation models. This role suits an experienced machine learning researcher and software engineer who can manage large-scale GPU clusters and collaborate on-site in Germany.

Role Overview

Aleph Alpha seeks a Senior AI Researcher to join the Pre-training team in Heidelberg. The person in this position will own critical technical areas like architecture, optimization, stability, and scaling for foundational models that serve sectors such as finance, manufacturing, and public administration. Working at the intersection of research and engineering, the professional will translate mathematical reasoning into principled training decisions across thousands of NVIDIA Blackwell GPUs.

Key Responsibilities

Manage core elements of the training recipe, including optimizers, schedules, and initialization, alongside PyTorch-based architectural improvements.
Create hyperparameter scaling laws and scale-up methodologies to predict multi-thousand-GPU behavior using small-scale proxy experiments.
Troubleshoot convergence issues, loss spikes, divergence, and distributed system failures like communication bottlenecks and race conditions.
Partner with Compute Performance, Data, Evaluation, and Post-Training teams to align the model lifecycle with hardware constraints and memory bandwidth.

Requirements

Proficiency in Python and deep familiarity with PyTorch-based training workflows.
Track record in machine learning research and software engineering through shipped models, open-source contributions, or publications.
Strong mathematical foundation in optimization, scaling behavior, and training dynamics.
Thorough understanding of transformer training dynamics and large distributed training jobs.
Ability to design rigorous experiments and translate empirical observations into robust training decisions.
Hands-on background pre-training large models with 7B+ parameters on substantial infrastructure like 100+ GPU clusters.
Adherence to strong software engineering practices for maintainable, tested, and reproducible code.
Capacity to work in Germany and collaborate regularly on site in Heidelberg.

Nice to Have

Hands-on experience training large language models or multimodal models on large GPU clusters using distributed frameworks such as Megatron-LM, DeepSpeed, or torchtitan.
Familiarity with predictive scaling, hyperparameter transfer, and scaling laws.
Experience profiling distributed jobs and diagnosing training anomalies like numerical instability.
Exposure to sparse training approaches including Mixture-of-Experts and their routing trade-offs.
Research excellence demonstrated through top-tier publications in venues like NeurIPS, ICML, or ICLR.
Curiosity regarding hardware and systems constraints affecting model scale.

Skills and Tools

Python
PyTorch
Megatron-LM
DeepSpeed
torchtitan
NVIDIA Blackwell GPUs

Salary and Benefits

30 days of paid vacation
Fitness and wellness offerings via Wellhub
Mental health support through nilo.health
Substantially subsidized company pension plan
Subsidized Germany-wide transportation ticket
Budget for additional technical equipment
Flexible working hours and hybrid working model
JobRad bike lease option

About the Company

Aleph Alpha operates in Europe, focusing on serious foundation model pre-training. The company builds models designed to understand German, comply with European regulatory requirements, and function reliably in high-stakes settings for customers in finance, manufacturing, and public administration. The organization maintains a flat structure, an emphasis on ownership and autonomy, and a collaborative team environment utilizing pair programming and mob programming.

How to Apply

Interested candidates should visit the official company careers page to submit an application.

Skills and tech stack

Questions candidates ask

Where is the job located?

The role is located in Heidelberg, Germany, with regular on-site collaboration required as part of the Pre-training team.

What qualifications are needed for the AI Researcher position?

Candidates must be proficient in Python and PyTorch, have a strong mathematical foundation in optimization and scaling, possess a track record in machine learning research and software engineering, understand transformer training dynamics, and have hands-on experience pre-training large models on substantial infrastructure.

What tools and distributed frameworks are mentioned?

The posting mentions Python, PyTorch, Megatron-LM, DeepSpeed, torchtitan, and NVIDIA Blackwell GPUs.

What benefits does Aleph Alpha offer?

Benefits include 30 days of paid vacation, Wellhub fitness access, nilo.health mental health support, a subsidized company pension plan, a subsidized Germany-wide transportation ticket, a technical equipment budget, flexible working hours with a hybrid model, and a JobRad bike lease.

Do I need to be able to work in Germany?

Yes, applicants must be able to work in Germany and collaborate regularly on site in Heidelberg.

Ready to apply?

The application is on the official AlephAlpha careers page.

Apply on company site