Machine Learning Engineer-Model Training Infrastructure

Responsibilities:

• Responsible for the design and implementation of a global-scale machine learning training system for feeds, ads and search ranking models.
• Responsible for the design and the implementation of orchestration layer of machine learning offline/online training processes.
• Responsible for improving use-ability and flexibility of the training APIs.
• Responsible for profiling and optimizing both training and validation frameworks to ensure efficient use of resources.
• Responsible for creating, managing, and optimizing data pipelines to ensure data availability for training.

Qualifications:

• Proficient in C/C++/CUDA/Python, and have solid programming skills.
• Familiar with deep learning frameworks (TensorFlow/Pytorch).
• Experience in developing and deploying large-scale systems.
• Ability to work independently and complete projects from beginning to end and in a timely manner.
• Good communication and teamwork skills to clearly communicate technical concepts with other teammates.
• Experience on improving core machine learning infrastructure (TensorFlow, Pytorch, and Jax).

Preferred Qualifications:

• Experience contributing to an open sourced machine learning framework (TensorFlow/PyTorch).
• Experience in big data frameworks (e.g., Spark/Hadoop/Flink), experience in resource management and task scheduling for large scale distributed systems.
• Strong background in one of the following fields: Hardware-Software Co-Design, High Performance Computing, ML Hardware Acceleration (e.g., GPU/RDMA) or ML for Systems.