Senior ML Engineer: High-Scale Inference & GPU Optimization

Jobgether; Málaga; Presencial; EUR 70.000 - 110.000; Jornada completa

Descripción de la vacante

Jobgether is listing on behalf of a partner company. A Senior Machine Learning Engineer (Token Factory) based in Spain will join a team building large-scale AI systems and high-performance infrastructure for training and serving foundation models at scale.

You will optimize inference, training pipelines, and GPU efficiency on massive GPU fleets, collaborating with ML, systems, and infrastructure experts in a fast-paced research environment.

Formación

• Strong ML fundamentals, transformer and LLM knowledge.
• Experience profiling and optimizing GPU workloads with Nsight or PyTorch Profiler.
• Deep GPU architecture understanding, memory hierarchy and compute vs memory trade-offs.
• Familiar with attention, KV-cache, Flash Attention, quantization.
• Experience with distributed training, sharding, and custom kernels.
• Proficient in Python and modern ML frameworks.
• Knowledge of CI/CD, version control, and testing.
• Excellent cross-functional collaboration.

Responsabilidades

• Drive inference optimization and latency reduction across diverse LLMs.
• Design and evolve inference engines, including KV-cache and dense/MoE models support.
• Develop low-precision training/inference pipelines (e.g., FP8) for large GPU clusters.
• Profile GPU workloads to identify bottlenecks and guide architecture decisions.
• Collaborate on scalable distributed training/inference systems with sharding and custom kernels.
• Contribute to engineering best practices: testing, CI/CD, production-grade ML systems.

Conocimientos

ML fundamentals

GPU profiling

GPU architecture

LLM concepts

Distributed training

Python programming

CI/CD

Communication skills

Team collaboration

Herramientas

Nsight

PyTorch Profiler

CUDA

Descripción del empleo

Jobgether is listing on behalf of a partner company. A Senior Machine Learning Engineer (Token Factory) based in Spain will join a team building large-scale AI systems and high-performance infrastructure for training and serving foundation models at scale.

You will optimize inference, training pipelines, and GPU efficiency on massive GPU fleets, collaborating with ML, systems, and infrastructure experts in a fast-paced research environment.