AI Research Scientist, SysML

Job Title

AI Research Scientist, SysML

Role Summary

Join Reality Labs' silicon and systems research team to develop and deploy efficient machine-learning models and system-level solutions for AR/VR and wearable devices. Work spans model algorithms, compression, numerics, hardware-aware design, and integration with silicon and system software.

The role focuses on enabling power- and compute-efficient training and on-device inference of vision and language models under real-time latency, power, and memory constraints.

Experience Level

Mid-level research scientist role. Expect experience designing or researching ML algorithms, system integration, or hardware acceleration for production or research projects.

Responsibilities

Key responsibilities include:

• Identify and solve cross-disciplinary ML acceleration problems spanning algorithms, model architecture, software, and hardware for AR/VR and wearable use cases.
• Invent and implement model compression and efficiency techniques (e.g., pruning, distillation, numerics, memory-efficient transformer architectures) and benchmark them across networks.
• Post-train and optimize foundation models to achieve Pareto-optimal trade-offs for multimodal LLMs and GenAI models on target hardware.
• Co-design hardware and system architectures with silicon teams and integrate algorithm and software changes to exploit hardware capabilities.
• Define use cases, evaluation methodology, and benchmarks; perform data-driven trade-off analysis and communicate recommendations to partners.
• Contribute to publications and patents on efficient on-device ML research.

Requirements

Must-have and nice-to-have qualifications.

• Must-have: Experience developing AI system infrastructure, AI algorithms, or AI hardware acceleration using C/C++ or Python.
• Must-have: Demonstrated ability to research and engineer efficient ML methods or systems (research publications, production impact, or equivalent experience).
• Nice-to-have: Experience with PyTorch, TensorFlow, or similar ML toolsets.
• Nice-to-have: Knowledge of architecting ML hardware accelerators and systems, or of on-device inference and hardware-aware model optimization.
• Nice-to-have: Experience optimizing post-training inference of large GenAI/LLM models under latency, power, and memory constraints.
• Nice-to-have: Track record of publications, patents, grants, or widely used open-source contributions; experience in responsible AI practices.

Education Requirements

PhD in Electrical Engineering, Computer Science, or equivalent experience is listed. The posting also references a Bachelor's in Computer Science, Computer Engineering, or relevant technical field or "equivalent practical experience." Fields mentioned: Computer Science, Computer Engineering, Electrical Engineering, and related technical fields. (Equivalent practical experience is acceptable.)

About the Company

Company: Meta Platforms

Headquarters: Menlo Park, California, United States

American technology company that develops social networking products (Facebook, Instagram, WhatsApp) and invests in virtual/augmented reality hardware and software through Reality Labs, focusing on connectivity, advertising, and immersive computing experiences.