Software Engineer III

Position:
Software Engineer III

Location:

Redmond, Washington Duration:
Contract Job  Job Overview:
This role is an onsite contract position in Redmond, focused on bringing state-of-the-art machine learning models from research into real-time deployment on embedded devices for AR/VR and wearable technologies. As a Software Engineer III, you will work within a Reality Labs-style research environment, collaborating closely with researchers and engineers to design, train, optimize, and deploy ML models that power next-generation audio and multimodal experiences for augmented reality glasses and related devices.

You will operate at the intersection of ML research, embedded systems, and high-performance C/C++ engineering, ensuring models run efficiently within strict latency, memory, compute, and power constraints. Responsibilities:
Collaborate side by side with research scientists to transition ML models from research prototypes into efficient, real-time inference solutions on embedded devices. Train, fine-tune, and adapt ML models with deployment constraints in mind, including latency, memory footprint, power consumption, and compute limitations. Optimize models for on-device inference using techniques such as quantization, pruning, distillation, operator fusion, and architecture refinements. Port, integrate, and profile ML models on embedded targets such as DSPs, NPUs, microcontrollers, and mobile SoCs, identifying and resolving performance bottlenecks.

Write clean, maintainable, and high-performance C/C++ code for on-device inference pipelines and integrate it into real-time prototype systems. Develop tools, scripts, and infrastructure that enable researchers to quickly evaluate model performance, accuracy, and resource usage on target hardware. Work within deep learning frameworks (e.g., PyTorch or Tensor Flow) to build, train, and iterate on ML models for audio, vision, and multimodal applications.

Coordinate with internal hardware, runtime, and toolchain teams to evaluate, integrate, and leverage existing systems, SDKs, and inference runtimes. Contribute to the design of end-to-end ML pipelines, from data preparation and training to deployment and monitoring on embedded platforms. Profile and debug performance issues across both the ML model and the underlying hardware, using appropriate tools and methodologies. Participate in code reviews, documentation, and knowledge sharing to maintain high engineering standards and foster team-wide best practices.

Work collaboratively as part of a multifunctional team where communication, documentation, and teamwork are highly valued. Continuously learn and explore new technologies, algorithms, and hardware capabilities relevant to AR/VR, audio, and embedded ML. Embrace ambiguity and contribute to problem solving at the frontiers of research, helping shape the future of AR/VR and wearable experiences.

Qualifications:

BS degree in Computer Science, Computer Engineering, Electrical Engineering, or a related field, or equivalent work experience. Industry experience optimizing and running ML inference on embedded or resource-constrained devices. Hands-on experience training ML models using deep learning frameworks such as PyTorch or Tensor Flow. 2+ years of industry experience writing production-quality C/C++ code. Experience profiling and debugging performance issues across ML models and target hardware platforms.

Experience working in cross-functional teams, collaborating with researchers, hardware engineers, and software developers. Strong understanding of on-device ML as a co-design problem, balancing model architecture, quantization, and hardware capabilities. Ability to write clean, efficient, and well-documented code suitable for real-time systems and embedded environments. Familiarity with embedded targets such as DSPs, NPUs, microcontrollers, and mobile SoCs, and their associated tool chains.

Preferred:

Experience with ML for audio applications (e.g., speech enhancement, source separation, keyword spotting, acoustic event detection). Preferred:

Experience with ML for computer vision (e.g., detection, tracking, segmentation, pose estimation). Preferred:
Knowledge of audio DSP fundamentals, including filtering, STFT/filter banks, beamforming, and real-time audio processing. Preferred:

Experience with on-device inference frameworks and tool chains such as Execu Torch, TFLite/LiteRT, ONNX Runtime, or vendor NPU/DSP SDKs. Preferred:

Experience with model compression techniques such…