Senior Distributed ML Engineer

European Tech Recruit; Berlin; Hybrid; EUR 80.000 - 110.000; Vollzeit

Zusammenfassung

European Tech Recruit is seeking a Senior Distributed ML Engineer for AI Safety Research in Berlin, operating in a hybrid setup. You will advance probabilistic inference methods for scalable, distributed ML training and inference, collaborating with researchers and engineers to accelerate development and deployment across distributed environments.

The role emphasizes profiling, optimization, and building tooling to streamline distributed workflows, leveraging Megatron, DeepSpeed, and related

Qualifikationen

• Master's or PhD preferred in CS/CE/SE.
• 3+ years designing distributed ML training systems.
• Hands-on with Megatron, DeepSpeed, Hugging Face Accelerate, or similar.
• Experience with cloud platforms (AWS, GCP, Azure) and containerization (Docker, Kubernetes).
• GPU profiling and optimization experience with PyTorch Profiler or NVIDIA Nsight.
• Strong problem-solving, debugging, and cross-team collaboration skills.

Aufgaben

• Collaborate with ML researchers and engineers to accelerate model development across distributed computing environments.
• Build, scale, and improve the reliability and efficiency of distributed training and inference workflows.
• Investigate performance bottlenecks across distributed ML workloads.
• Profile code and optimise GPU, CPU, memory, and networking resources.
• Develop tools and orchestration to enable access to distributed computing resources.

Kenntnisse

Problem solving

Communication

Cross-team collaboration

GPU profiling experience

Ausbildung

Master's degree or PhD preferred in CS/CE/SE

Tools

Megatron

DeepSpeed

Hugging Face Accelerate

FSDP

vLLM

verl

PyTorch Profiler

NVIDIA Nsight

Ray

SLURM

Docker

Kubernetes

gRPC

vector databases

Jobbeschreibung

Senior Distributed ML Engineer - AI Safety Research Non-Profit - Berlin (Hybrid)

We are seeking a Machine Learning (ML) Research Scientist specialising in probabilistic inference to join a highly technical research team working at the forefront of machine learning and probabilistic modelling. In this role, you will develop and evaluate advanced probabilistic inference methods, with a particular focus on amortised inference, translating theoretical insights into practical, scalable implementations.

• Collaborate with ML researchers and engineers to accelerate model development, training, and inference across distributed computing environments.
• Build, scale, and improve the reliability and efficiency of distributed training and inference workflows.

System Optimization & Bottleneck Diagnosis:

• Investigate performance bottlenecks across distributed ML workloads.
• Profile code and optimise the utilisation of GPU, CPU, memory, and networking resources.

Tooling & Best Practices:

• Develop tools, libraries, and orchestration infrastructure to make distributed computing resources accessible.
• Establish, document, and maintain engineering best practices across technical teams while evaluating emerging technologies.

Requirements:

• Degree in Computer Science, Computer Engineering, Software Engineering, or a related field (Master's or PhD preferred).
• 3+ years of experience designing and implementing distributed machine learning training systems or frameworks.
• Hands-on experience with technologies like Megatron, DeepSpeed, Hugging Face Accelerate, FSDP, vLLM, or verl.
• Experience with cloud platforms (AWS, GCP, Azure), workload managers (Ray, SLURM), containerization (Docker, Kubernetes), gRPC, and vector databases.
• Proven track record in GPU performance profiling and optimisation using tools like PyTorch Profiler, PyProf, or NVIDIA Nsight.
• Strong problem-solving, debugging, communication, and cross-team collaboration skills.