Senior AI R&D Engineer (Model Training)

Aleph Alpha GmbH; Deutschland; On-site; EUR 90,000 - 150,000; Full time

Benefits offered by this job

30 Days vacation

Flexible working hours

Work with international industry and科学

Join a dynamic start-up and rapidly-g>

Job summary

Aleph Alpha GmbH is seeking a Senior AI R&D Engineer in Model Training to own and evolve the end-to-end training pipeline. You will optimize scheduling, enable large-scale code execution for RL, and lift inference throughput while building high-quality tooling and infrastructure.

You will work across disciplines, drive ownership, and shape future architecture. Join a fast-moving team in Germany with opportunities to relocate to Heidelberg or Berlin, and collaborate with international experts on

Qualifications

• Degree in computer science, engineering, or related field.
• Strong collaboration and communication with diverse audiences.
• Willingness to relocate to Germany; Heidelberg preferred, Berlin also possible.
• Experience contributing in high-performing teams and ownership of problems.
• Experience with distributed systems and Kubernetes.
• Interest in state-of-the-art foundation models and ML learning.

Responsibilities

• Co-own the training pipeline end-to-end and maintain infrastructure for experiments.
• Optimize training schedules and improve throughput of evaluation and inference.
• Develop tooling to speed up experimentation and model iterations.
• Collaborate across disciplines; learn from researchers and engineers.
• Drive architectural improvements and ownership of infrastructure decisions.

Skills

Foundation models

High-performing teams

Ownership mindset

Strong communication

Kubernetes

Distributed systems

Relocation to Germany

ML interest

Education

Degree in computer science or engineering

Tools

Kubernetes

Job description

• As a Senior AI R&D Engineer in Model Training (f/m/d), you’ll work across our full stack
• Some weeks you might be optimizing how training loads are scheduled on our cluster and making the pipeline more robust and performant so we can iterate faster
• Other weeks, you’ll be enabling large-scale code execution for reinforcement learning
• And at other times, you might dig deep into our evaluation codebase to lift inference throughput on evals
• No two days are the same. Things move fast, and your ability to focus and prioritize is what lets you unblock the team day-to-day while designing the high-quality tooling and infrastructure that speeds us up long-term
• We’re still building out our training pipeline and infrastructure. Some pieces exist, some don’t, and you’ll have real influence on what gets built and how
• Your work directly shapes how quickly we can experiment and improve our models
• Co-own the training pipeline end-to-end. Design, build, and maintain the infrastructure and components that let us iterate fast on experiments
• Build high-quality tooling. Model training is a continuous effort, and we deliberately invest in our tooling and infrastructure to stay successful long term
• Collaborate across disciplines. We believe in cross-functional teams. Engineers and researchers work closely so we can learn from each other and iterate faster together
• Champion good engineering practices. Working incrementally, maintaining fast feedback loops, and refactoring continuously keep a team successful long-term, especially when moving fast
• Shape the direction of the team. Our culture empowers individuals to take ownership. If you see that we’ll need more GPUs, a different storage system, or a change to how the team is set up, you should drive this change

Benefits

• 30 Days of paid vacation
• Flexible working hours
• Work with international industry and science experts
• Join a dynamic start-up and a rapidly growing team
• Take on responsibility and shape our company and technology
• Regular team events

Requirements

• Deep interest in how state-of-the-art foundation models work
• Experience contributing in high-performing teams
• Desire to take ownership of problems and collaborate with other teams to solve them
• Strong communication skills, with the ability to convey technical solutions to diverse audiences
• Degree in computer science, engineering, or a related field
• Ability to write software that other strong engineers want to read and build on
• Willingness to relocate to Germany. Our primary working locations are Heidelberg (preferred) and Berlin, although there is some flexibility to work from other locations in Germany with regular travel to Heidelberg (potentially weekly)
• A track record of taking initiative to deliver high-impact work
• Experience working with Kubernetes
• Experience working with distributed systems
• We do not require prior experience in machine learning for this role, but we do value your eagerness to learn. If you have prior experience in ML, we will be particularly excited about:
  • Experience bringing AI research innovations into production
  • Experience in areas such as large-scale data processing or distributed computation for foundation model training or inference
  • Experience with performance engineering: profiling, benchmarking, and optimizing code for throughput, latency, or memory