Senior AI Infrastructure & Frontier Models Architect

BMW Group; New Germany (MN); Presencial; USD 150.000 - 230.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Vacation pay

Christmas bonus

Profit sharing

Flexible working hours

6 weeks annual leave

Discounted BMW & MINI

Descripción de la vacante

BMW Group is hiring a Senior AI Infrastructure & Frontier Model Engineer to design and operate HPC/GPU clusters enabling training, fine-tuning, and inference of advanced AI models. You will integrate frontier models into production environments and continuously improve model quality and performance.

The role requires a university degree in CS/EE and 3+ years in ML/AI with hands-on experience in PyTorch, DeepSpeed, LoRA, RLHF, and distributed training.

Formación

• University degree in Computer Science, Electrical Engineering, or equivalent.
• 3+ years in ML/AI and LLMs.
• Experience building HPC/GPU AI platforms.
• Proficient with PyTorch, DeepSpeed, LoRA, RLHF and evaluating large-scale models.

Responsabilidades

• Operate HPC/GPU clusters for AI workloads.
• Support training, fine-tuning, and inference of AI models.
• Integrate frontier models into development and production environments.
• Optimize performance, scalability, and resource efficiency of AI infra.
• Collaborate with cross-functional teams to advance AI platforms.

Conocimientos

Machine Learning

AI Models

Distributed Training

Linux

Kubernetes

Docker

PyTorch

DeepSpeed

Megatron-LM

LoRA

QLoRA

RLHF

DPO

Frontier Models

Educación

University degree

Herramientas

HPC

GPU Clusters

Descripción del empleo

BMW Group is hiring a Senior AI Infrastructure & Frontier Model Engineer to design and operate HPC/GPU clusters enabling training, fine-tuning, and inference of advanced AI models. You will integrate frontier models into production environments and continuously improve model quality and performance.

The role requires a university degree in CS/EE and 3+ years in ML/AI with hands-on experience in PyTorch, DeepSpeed, LoRA, RLHF, and distributed training.