C++/CUDA Systems Engineer – Surgical Robotics Platform

Horizon Surgical Systems, Inc.•Los Angeles, CA

8h•$140,000 - $160,000

About The Position

Horizon Surgical Systems Inc. is revolutionizing the world of surgical ophthalmology by developing a novel, AI driven, and imaging-guided surgical robotic system. Horizon Surgical Systems Inc. aims to expand access to care, provide superior capabilities to the human surgeon, and enhance patient outcomes. Microsurgery in general and Ophthalmology are subfields of surgery for which the surgical outcomes can be significantly improved by robotic systems to allow superior dexterity, precision, accuracy, and visualization beyond the human surgeon’s own capabilities. We are seeking highly motivated, and intellectually inquisitive individuals looking to make a positive impact on healthcare via the development of robotic technology. The core values of Horizon Surgical Systems Inc. are: Commitment to Excellence: We aim to deliver superior patient outcomes and surgeon experiences Passion for Creativity and Innovation: We are driven by new ideas and aim to push the boundaries of what's possible Teamwork and Camaraderie: We achieve our best when we collaborate and work together Welcoming of Critical Opinion: We are enriched by constructive criticism and support the best ideas Personal Accountability: We honor our commitments and take responsibility for our actions Horizon Surgical Systems Inc. offers: An opportunity to build autonomous surgical robotic systems driven by image guidance and AI technology for the future of affordable, high-quality healthcare. The opportunity to work alongside clinicians, engineers, and global leaders in cutting-edge AI, imaging, and robotics technology. Competitive compensation and an excellent company-paid benefits package. Role: Horizon Surgical Systems is building Polaris, the first autonomous cataract surgery platform. You will work on a C++/CUDA-based runtime for real-time surgical imaging and perception, operating in a live surgical environment where latency, reliability, and determinism are critical. This role sits within the AI team but focuses on production systems: GPU inference pipelines, device orchestration, and system integration. You will own components that span from hardware-facing interfaces to high-performance compute pipelines, and ensure they operate cohesively under real-time constraints.

Requirements

Bachelor’s or Master’s degree in Computer Science, Robotics, Electrical Engineering, or related field
3+ years of experience writing production C++ in systems, robotics, or embedded environments
Strong C++17/20 fundamentals (multithreading, memory management, concurrency patterns)
Experience with CUDA or GPU-accelerated compute pipelines
Experience building or maintaining real-time or latency-constrained systems
Familiarity with logging, diagnostics, and performance profiling in production environments

Nice To Haves

Experience in medical devices, robotics, or other safety-critical systems
Familiarity with ML inference frameworks (e.g., TensorRT, ONNX Runtime)
Experience integrating and debugging physical hardware (cameras, motors, sensors)
Familiarity with inter-process communication (DDS, shared memory, or similar)
Experience with performance profiling tools (Nsight Systems, perf, nvprof, or equivalent)

Responsibilities

Design and maintain multi-threaded C++ pipelines for real-time model inference, integrated with robotic control and perception systems
Implement scheduling and resource management across CPU/GPU to meet strict latency and throughput constraints
Develop infrastructure for coordinating physical devices (cameras, sensors, motion systems), including sequencing, configuration, and fault handling
Define and implement system interfaces (APIs, messaging, IPC) to integrate subsystems into a coherent runtime
Build and maintain automated testing at component and system levels, including simulation and hardware-in-the-loop where applicable
Contribute to profiling, observability, and debugging tooling for performance-critical systems