Director, Model Inference Engineering
Core
Lead and scale the Mariner inference platform serving Thomson Reuters' custom LLM family for professional workflows, ensuring production footing, cost-efficiency, and compliance.
Role type
Director, Model Inference Engineering
Builds
Mariner (custom inference platform) and Thomson LLM model family serving production calls
Domain
Legal/Financial technology, Large Language Models (LLM), Cloud Infrastructure
Deliverable
production ML models
Required skills
Production ML-serving leadership, Infrastructure cost ownership, Team building and scaling, Resource constraint management, External-facing deployment experience, Operational excellence
Preferred skills
LLM-serving stacks (vLLM, TensorRT-LLM, Triton, SGLang), Research-product roadmap alignment, GPU-cluster capacity planning, Systems languages (Rust, Go), Regulated environment experience
Technologies
Kubernetes, vLLM, TensorRT-LLM, Triton, SGLang, Rust, Go
Responsibilities
Grow the inference-engineering team, Own delivery of the engineering roadmap, Scale production footing and compliance, Represent tradeoffs to leadership, Produce token-economics benchmarks
Seniority
Director, strategic leadership with hands-on technical ownership