Sr. Principal Software Engineer, AI-Video Technical Lead
Owned the AI-video platform for medical surgical-guidance device, from GPU scheduling to release strategy.
Impact
- Architected a regulated AI-video platform with three core goals: isolate multiple AI apps sharing one GPU, enforce clear medical/non-medical boundaries, and low overlay latency for real-time surgical guidance.
- Designed core runtime capabilities including AI scheduling, inference hardening, video and AI data recording, overlay generation, time synchronisation, and benchmarking.
- Act as the technical bridge across ML, Systems, Quality, Regulatory, and UI, turning requirements into a delivery roadmap.
- Built the platform to scale future AI features through shared x86/ARM dev containers, a modular build system, common libraries, and automated end-to-end model verification before deployment.
- Designed hardware-in-the-loop testing, CI-published documentation, and variant-aware build/versioning practices to make releases, regression testing, and submission artifacts repeatable under change control.
- Technical lead for a team of 5 engineers, combining hands-on architecture work with mentoring, reviews, and pairing.
Role notes ▸
Promoted to Sr. Principal Software Engineer and AI-Video Technical Lead to own the architecture and roadmap for Medtronic's surgical AI-video platform.
Led the move from prototype to release-ready system by designing the GPU scheduler, inference runtime, safety-overlay path, and recording/logging pipeline; driving FDA-submission work for AI-video features; and leading a team of 5 across ML, systems, quality, regulatory, and UI.