Build production AI inference systems at Thomson Reuters using GPU optimization and Kubernetes across a flexible hybrid working setting. Support reliable services for AI-driven products and enterprise workflows. As a Staff AI Inference Engineer, you will productionize, optimize, and scale classical machine learning, generative AI, and large language model workloads. You will collaborate with Platform Engineering, Product, Data Science, Architecture, Enterprise AI, and cloud engineering teams. The work includes capacity forecasting, model onboarding, production APIs, and multi-cloud infrastructure across AWS, Azure, Google Cloud Platform, and Oracle Cloud Infrastructure. Key Responsibilities:

Optimize models through quantization, pruning, distillation,



and hardware-specific tuning Deploy GPU inference workloads across multi-cloud and Kubernetes environments Implement routing and failover for OpenAI, Anthropic, and Vertex AI traffic Build containerized inference pipelines and production-grade APIs Monitor bottlenecks, latency, utilization, reliability, and model drift Requirements:

Seven or more years of relevant experience Strong machine learning, large language model, and inference optimization fundamentals GPU programming experience; CUDA preferred TensorRT, ONNX Runtime, PyTorch, or TensorFlow experience Python plus a systems language; C++ strongly preferred Apply staff-level inference expertise across GPU systems, Kubernetes, multi-cloud platforms, and enterprise AI reliability at Thomson Reuters.

📌 Staff AI Inference Engineer (Toronto)
🏢 Thomson Reuters
📍 Toronto