Senior AI R&D Engineer – Model Training (f/m/d)

The role

You will build and improve the model training pipeline used to train Kolibri, working across scheduling, reinforcement learning code execution and evaluation code. You will work mostly in mob or pair programming with your team.

What you will do

• Optimize how training loads are scheduled on the cluster and make the pipeline more robust and performant.
• Enable large-scale code execution for reinforcement learning.
• Improve inference throughput on evaluation code.
• Improve maintainability of code through tests and design and architecture changes.
• Work with your team on design decisions and priorities.
• Identify issues and take initiative.

Required qualifications

• Ability to dive into a codebase, change its behavior, and leave it better tested and easier to understand.
• Understanding of designing for low coupling and high cohesion.
• Ability to refactor incrementally.

Nice to have

• Experience with Python, the main language used.
• Experience with Rust, used occasionally.
• Experience with Kubernetes, which much of the training pipeline is built on.
• Experience with Test Driven Development (TDD), trunk based development, continuous integration and Domain Driven Design.

Benefits

• 30 days of paid vacation.
• Access to 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 a hybrid working model.
• JobRad bike lease.

Before you apply

• The posting says the company hires for ability to learn and mindset rather than specific technology skills.

About the company

Aleph Alpha builds foundation models in Europe, including the recently released Kolibri-1, for customers in finance, manufacturing and public administration. The team is based in Heidelberg and values ownership, autonomy and a flat organizational structure.