Latest AI JobsBrowse Jobs
pragmatike logo

AI Infrastructure Engineer (GPU) - Remote EMEA

pragmatikeUkraineWork with our Clients
Apply on company site
Salary
Not listed
Job type
Full-time
Work mode
Remote
Experience
Senior · 4+ yrs
Posted
23 Sept 2026
Status
Open · checked 7d ago

About this role

Pragmatike is hiring a remote AI Infrastructure Engineer in the EMEA timezone to build and operate large-scale machine learning inference platforms. This role suits infrastructure professionals with production-grade model serving experience who want to design decentralized GPU compute systems for a well-funded European cloud startup.

Role Overview

Pragmatike is recruiting on behalf of a distributed cloud infrastructure startup that provides GPU-powered services for artificial intelligence workloads. The position focuses on developing reliable, scalable, and low-latency machine learning inference platforms using decentralized architecture.

Key Responsibilities

Build and operate production-grade model serving infrastructure using tools like vLLM, Triton, or TGI.
Design and implement deployment pipelines featuring canary and blue/green rollout strategies.
Develop multi-model serving architectures, auto-scaling systems, and request routing layers.
Optimize GPU memory efficiency, utilization, network throughput, and storage performance.
Construct observability systems to monitor inference latency, throughput, hardware metrics, and system health.
Manage CI/CD pipelines and model registries for reproducible deployments.
Oversee the complete lifecycle of machine learning systems, including on-call duties and operational support.
Establish engineering best practices within a fast-moving startup environment.

Requirements

Four or more years of professional background in platform engineering, ML Ops, site reliability engineering, or related infrastructure domains.
Practical background utilizing model serving frameworks such as Triton, TGI, or vLLM.
Container orchestration expertise managing GPU workloads in production environments.
Familiarity with MLOps tooling, including automated deployment pipelines, experiment tracking, and model registries.
Programming capability in Python alongside infrastructure-as-code tools like Terraform or Helm.
Knowledge of performance tuning, distributed systems, and production reliability.
Capability to leverage AI coding assistants for debugging and development workflows.
Independent working style suited for a remote-first setting.

Nice to Have

Background with machine learning platforms including MLflow, Kubeflow, or KubeAI.
Understanding of multi-tenant inference, GPU scheduling, or ROCm and CUDA optimization.
Experience optimizing costs across various inference workloads and GPU types.
Previous tenure in early-stage startups or greenfield infrastructure initiatives.
History of constructing production systems from the ground up rather than supporting legacy software.

Skills and Tools

Python
Terraform
Helm
vLLM
TGI
Triton
Kubeflow
MLflow
KubeAI
CUDA
ROCm

About the Company

Pragmatike acts as a recruiter for a well-funded, fast-scaling distributed cloud infrastructure enterprise delivering GPU-powered resources for artificial intelligence applications. The organization focuses on secure storage, high-speed data transfer, and lower environmental impact than legacy cloud providers.

How to Apply

Interested candidates should apply on the official company careers page.

Skills and tech stack

PythonTerraformHelmvLLMTGITritonKubeflowMLflowKubeAICUDAROCmMLOpsMachine LearningDistributed Systems

Questions candidates ask

Where is the AI Infrastructure Engineer position located?

The role is fully remote within the EMEA timezone, with Pragmatike recruiting on behalf of the client.

How many years of experience are required for this role?

Candidates must have four or more years of experience in ML Ops, Platform Engineering, SRE, or related infrastructure roles focused on machine learning systems.

What programming language and tools are required?

Proficiency in Python is required, alongside infrastructure-as-code tools such as Terraform or Helm, and model serving frameworks like vLLM, TGI, or Triton.

What languages must applicants speak?

Fluent English is required for the position.

What are the primary responsibilities of the engineer?

Responsibilities include building production-grade model serving infrastructure, designing deployment pipelines, optimizing GPU utilization, and developing auto-scaling systems.

Ready to apply?

The application is on the official pragmatike careers page.

Apply on company site