Senior Agentic AI Performance Optimizer — Remote

Advanced Micro Devices; Stockholms kommun; On-site; SEK 1,011,000 - 1,348,000; Full time

Job summary

Advanced Micro Devices (AMD) is seeking a Principal Agentic AI Performance Optimization Engineer in Europe. The role focuses on end-to-end optimization of AI workloads on AMD hardware, combining platform engineering with hands-on inference serving and agentic systems.

You will develop tooling, safety frameworks, and abstractions to support diverse workloads and hardware platforms while packaging validated optimizations for deployment. Remote-friendly across Europe.

Qualifications

• Experience building developer tooling, command line tools, or automation platforms in Python.
• Experience deploying LLMs or AI models in production with frameworks like vLLM or SGLang.
• Familiarity with agentic systems and AI-driven workflows.
• Demonstrated success designing scalable software solutions.
• Experience with Docker and cluster scheduling platforms.
• Degree in Computer Science or related field, or equivalent practical experience.

Responsibilities

• Develop and enhance tooling across the performance optimization stack.
• Create safety and validation frameworks for agentic systems.
• Design abstractions supporting multiple workloads and platforms.
• Package optimizations into deployable artifacts with provenance.
• Drive adoption of shared tooling across teams and repositories.

Skills

Python

LLM deployment

Agentic systems

CI/CD

Docker

Kubernetes

C/C++

CUDA

Observability

GPU architecture

Education

Bachelor's degree in CS or related field

Tools

vLLM

SGLang

Helm

Docker

Kubernetes

Job description

Advanced Micro Devices (AMD) is seeking a Principal Agentic AI Performance Optimization Engineer in Europe. The role focuses on end-to-end optimization of AI workloads on AMD hardware, combining platform engineering with hands-on inference serving and agentic systems.

You will develop tooling, safety frameworks, and abstractions to support diverse workloads and hardware platforms while packaging validated optimizations for deployment. Remote-friendly across Europe.