Senior AI Inference Engineer - Hybrid & Cloud-Scale

Thomson Reuters; Eagan (MN); Hybrid; USD 127,000 - 237,000; Full time

Benefits offered by this job

Mental Health Days

Headspace app

401k with company match

Tuition Reimbursement

Volunteer days off

Job summary

Thomson Reuters seeks a Staff Inference Engineer, AI to productionize and optimize AI and LLM workloads across a multi-cloud footprint. You will deploy and scale inference using GPUs, build containerized pipelines on Kubernetes, and collaborate with Platform Engineering to ensure reliability and governance across enterprise AI services.

The role requires strong ML/LLM fundamentals, CUDA, TensorRT/ONNXRuntime, and Python with C++ for performance-critical paths. Hybrid work in the US is available.

Qualifications

• 7+ years of ML/LLM experience with inference optimization.
• Hands-on GPU programming (CUDA) and inference runtimes (TensorRT, ONNXRuntime).
• Proficiency in Python and systems languages (C++ preferred).
• Experience deploying AI workloads to cloud and Kubernetes.

Responsibilities

• Productionize, optimize, and scale AI/LLM inference workloads.
• Deploy inference on GPUs across AWS/Azure/GCP/Kubernetes clusters.
• Implement routing and guardrails for model traffic and reliability.
• Collaborate with Platform Engineering to align cloud-native patterns.
• Develop containerized pipelines and monitor latency and bottlenecks.

Skills

ML fundamentals

Inference optimization

GPU programming

Python

C++

TensorRT

ONNX Runtime

PyTorch

Kubernetes

Job description

Thomson Reuters seeks a Staff Inference Engineer, AI to productionize and optimize AI and LLM workloads across a multi-cloud footprint. You will deploy and scale inference using GPUs, build containerized pipelines on Kubernetes, and collaborate with Platform Engineering to ensure reliability and governance across enterprise AI services.

The role requires strong ML/LLM fundamentals, CUDA, TensorRT/ONNXRuntime, and Python with C++ for performance-critical paths. Hybrid work in the US is available.