Build and optimize large-scale machine learning training and inference systems for real-time trading applications. Responsibilities include distributed training pipelines, experiment automation, GPU and low-level performance optimization, low-latency inference, scalable model infrastructure, framework improvements, and evaluation of ML tooling. The role requires production software engineering, strong ML fundamentals, systems thinking, performance measurement, and collaboration with quantitative researchers and engineering specialists.

Architect and co-design machine learning models for custom hardware, accounting for latency, resource, and numerical constraints. Translate model requirements into hardware architecture decisions, collaborate with hardware engineers to implement and verify inference solutions, and deploy systems from proof of concept through production. Evaluate research in neural architecture search, ML systems, quantization, and compiler infrastructure for measurable performance improvements.

Own end-to-end technical onboarding and certification for electronic trading clients, from connectivity testing through production launch. Coordinate Development, Infrastructure, Compliance, Legal, Clearing, Business Development, and clients; validate FIX connectivity and routing; configure accounts and trading setups; monitor post-launch message quality and behavior; manage readiness, blockers, recertification, regulatory reporting, and post-trade requirements. Serve as the primary responder for onboarding issues and ensure repeatable, operationally disciplined processes.