Robotics Research Engineer: From Prototypes to Production

XDOF; San Mateo (CA); On-site; USD 140,000 - 190,000; Full time

Job summary

XDOF is hiring a Research Engineer to bridge research prototypes and production-scale software for embedded robotics. You will convert experimental perception pipelines into reliable, real-time code running on hardware, while profiling and optimizing for CPU, memory and GPU efficiency.

The role spans working across perception, ML and data infrastructure, shipping production-grade systems and collaborating with researchers on tradeoffs between accuracy and latency.

Qualifications

• 3+ years of industry software engineering focused on systems, performance or production ML.
• Strong C++ proficiency with modern standards and memory management.
• CUDA programming: write, profile and debug GPU kernels.
• CPU performance optimization: profiling, cache behavior and SIMD.
• Python with ML frameworks to read/modify research code.
• Linux systems, build tools, debugging and containerization.

Responsibilities

• Transform research prototype code into production-grade, scalable software.
• Hardening perception pipelines for real-time embedded platforms.
• Profile and optimize performance at CPU, memory and GPU levels.
• Develop and debug CUDA kernels for acceleration.
• Integrate outputs into production codebase with testing and observability.
• Containerize workloads with Docker for scalable deployment.
• Collaborate with researchers to balance accuracy, latency and resources.

Skills

C++ (modern)

CUDA

Python

Linux

Performance optimization

Containerization

ML frameworks (PyTorch, TensorFlow)

Tools

NSight

Docker

Job description

XDOF is hiring a Research Engineer to bridge research prototypes and production-scale software for embedded robotics. You will convert experimental perception pipelines into reliable, real-time code running on hardware, while profiling and optimizing for CPU, memory and GPU efficiency.

The role spans working across perception, ML and data infrastructure, shipping production-grade systems and collaborating with researchers on tradeoffs between accuracy and latency.