Senior AI Inference Engineer - High-Throughput LLM Serving

Confidential; Ireland; On-site; EUR 120,000 - 180,000; Full time

Job summary

Confidential in Ireland is seeking a hands-on ML engineering lead to optimise LLM inference pipelines and own end-to-end model serving in production. You will deploy multi-GPU inference, manage low-latency pathways, and build OpenAI-compatible APIs across cloud and on-prem environments.

You should have deep experience with vLLM/TensorRT-LLM/TGI, strong GPU performance skills, and solid Python/GoLang software engineering.

Qualifications

• Hands-on production experience serving LLMs at scale with measurable throughput/latency improvements.
• Deep familiarity with a modern inference/serving framework (vLLM, TensorRT-LLM, TGI, or similar).
• Strong grip on GPU performance: memory management and model/tensor parallelism.
• Solid software engineering in Python or GoLang, plus Docker + Kubernetes for production deployment.
• A benchmarking mindset—measure, compare, and defend trade-offs.

Responsibilities

• Optimise LLM inference pipelines (multi-GPU inference, prefix caching, memory-efficient serving).
• Own end-to-end model serving in production (deployment, low-latency inference, multi-GPU parallelism).
• Build and maintain OpenAI-compatible serving APIs (/v1/chat, /v1/responses).
• Tune and operate a modern serving stack (vLLM) with batching and cache management.
• Maximise GPU usage across architectures; profile and remove bottlenecks.
• Instrument the serving layer with logging, telemetry, and metrics for observability and autoscaling.
• Ship on Kubernetes: Docker, Helm, CI/CD, staged rollouts.
• Benchmark rigorously against standards and build tooling for performance characterization.

Skills

Production experience

LLM inference

GPU performance

Python

Go

Docker

Kubernetes

Benchmarking

Tools

vLLM

TensorRT-LLM

TGI

Job description

Confidential in Ireland is seeking a hands-on ML engineering lead to optimise LLM inference pipelines and own end-to-end model serving in production. You will deploy multi-GPU inference, manage low-latency pathways, and build OpenAI-compatible APIs across cloud and on-prem environments.

You should have deep experience with vLLM/TensorRT-LLM/TGI, strong GPU performance skills, and solid Python/GoLang software engineering.