Staff ML Engineer, Ads ML Infrastructure

What you'll do

• Design and develop secure, scalable back-end systems for large-scale ML serving.
• Support real-time and batch inference using distributed systems like Ray and high-throughput RPCs.
• Optimize model performance through quantization, tensor parallelism, and inference acceleration techniques.
• Design and optimize low-level GPU kernels to maximize hardware utilization and accelerate deep learning primitives.
• Build foundational AI/ML platforms including Feature Stores and Vector Databases.
• Develop abstractions, automation tools, and reusable components to streamline ML workflows.
• Implement privacy-preserving ML techniques and federated learning to meet Apple's privacy commitments.
• Lead technical growth through code reviews, design reviews, and internal knowledge sharing.

What we're looking for

• Proven track record of designing and operating large-scale, low-latency ML Serving platforms for real-time and batch inference.
• Experience with model quantization, tensor parallelism, and inference optimizations like ONNX Runtime, TensorRT, or vLLM.
• Experience working on distributed systems such as Ray or high-throughput RPC systems for scalable inference workloads.
• Hands-on experience designing and optimizing low-level GPU kernels to maximize hardware utilization and accelerate deep learning primitives.
• Prior experience in the advertising industry, federated learning, and privacy-preserving ML techniques.
• Recognized technical leader capable of mentoring engineers through code reviews, working groups, and knowledge sharing.
• Experience leading development of foundational AI/ML platforms including Feature Stores and Vector DBs.
• PhD, MS, or BS in computer science or a related field with 8+ years of experience in machine learning (preferred).