Staff AI Inference Engineer: Multi-Cloud, LLMs & GPUs

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

Benefits offered by this job

Hybrid Work Model

Mental Health Days

Tuition Reimbursement

Employee Stock Purchase Plan

Job summary

Thomson Reuters in the United States (Eagan, MN) is seeking a Staff Inference Engineer, AI to productionize and scale AI/LLM workloads across AWS, Azure, GCP, and OCI. You will collaborate with platform engineering, data science, and product teams to onboard research models into production and optimize inference latency.

Responsibilities include deploying GPU-accelerated pipelines, implementing routing and failover, and ensuring observability and governance in line with TR AI standards.

Qualifications

• 7+ years of experience in ML/LLM or AI engineering.
• Strong understanding of ML inference optimization and deployment.
• Hands-on experience with CUDA, TensorRT/ONNX Runtime, and deep learning frameworks (PyTorch/TensorFlow).
• Proficiency in Python and at least one systems language (C++ preferred).
• Experience deploying AI workloads to AWS/GCP/Azure and Kubernetes.

Responsibilities

• Productionize and scale AI/LLM workloads on multi-cloud platforms.
• Optimize inference with quantization, pruning, and distillation.
• Deploy GPU-driven workloads on Kubernetes across AWS/Azure/GCP/OCI.
• Implement routing and failover for AI traffic.
• Integrate models into production-grade APIs and ensure observability.
• Collaborate with Platform Engineering and AI Services to meet reliability standards.

Skills

GPU programming

Python

C++ proficiency

PyTorch

TensorRT

ONNX Runtime

Kubernetes

Distributed systems

Tools

Snowflake

OpenSearch

API management

Job description

Thomson Reuters in the United States (Eagan, MN) is seeking a Staff Inference Engineer, AI to productionize and scale AI/LLM workloads across AWS, Azure, GCP, and OCI. You will collaborate with platform engineering, data science, and product teams to onboard research models into production and optimize inference latency.

Responsibilities include deploying GPU-accelerated pipelines, implementing routing and failover, and ensuring observability and governance in line with TR AI standards.