Machine Learning Engineer

DeepMotion is seeking a Machine Learning Engineer to develop the server-side capabilities that enable interactive AI-based motion capture and motion synthesis. The position centers on machine learning model development and deployment, along with inference framework engineering for a cloud-based animation creation platform.

Responsibilities

• Develop and train algorithms and models that power the motion perception and motion generation engine
• Design, build, and manage large-scale text, video, and motion datasets used to train multi-model motion perception and generation systems
• Develop, deploy, and optimize inference frameworks for motion perception and generation running in the cloud
• Research, analyze, develop, and test machine learning components that support business strategies and the product roadmap

Requirements

• BS or MS required in Computer Science, Engineering, or a closely related field, with specialization in computer vision, machine learning, robotics, or artificial intelligence
• Solid knowledge of at least one machine learning research framework, for example PyTorch or Tensorflow
• Solid knowledge of at least one high-performance inference framework, for example TensorRT or Apache TVM
• Experience profiling and optimizing deep neural networks, including familiarity with GPU profiling tools such as NVIDIA Nsight

Location

San Francisco Bay Area, CA (onsite)

Technologies

• PyTorch, Tensorflow
• TensorRT, Apache TVM
• NVIDIA Nsight
• OpenCV, PyAV
• Kubernetes
• AWS, GCP, Azure
• Python

Nice to Have

• Familiarity with Python-based image and video manipulation, including encoding and decoding frameworks such as OpenCV and PyAV
• Experience with cloud orchestration systems such as Kubernetes and cloud providers such as AWS, GCP, and Azure
• Knowledge of transformers, diffusion models, and multimodal discriminative and generative models
• Ability to write robust and maintainable client-server architectures and APIs