Peloton is looking for a Machine Learning Engineer, Deployment focused on Deep Learning/Computer Vision. You’ll be developing cutting-edge systems to provide our members with a world-class fitness experience, in collaboration with Product, Software, and Hardware teams.
- Own the deployment loop between research-driven AI models and their use in edge environments, with a heavy focus on performance and efficiency running on embedded, resource constrained hardware.
- Collaborate and work closely with engineers to translate and deploy new AI/ML solutions for connected fitness devices.
- Be the voice in the room that guides development work by ensuring work being done by the team is deployable in an end to end system.
- Ensure model performance remains within expected bounds when promoting experimental models to production.
- Specifically, you may encounter projects focused on: Temporal modeling, Object Detection, Segmentation, Perception, Multi-modal and Ensembling
- 2-5+ years of hands-on, real-world experience with one or more of Computer Vision, Machine Learning, Deep Learning.
- Experience with Qualcomm SNPE, Tensorflow Lite, or other similar Edge Inference/NN Acceleration frameworks.
- Experience with deploying AI models on the edge, including Model Compression techniques such as Quantization, Pruning, Distillation
- Proficiency in Python and Java/C/C++, and ML frameworks like PyTorch, Tensorflow, Keras, etc.
- Ability to quickly translate research work into high-quality production code with a strong sense of good system design.
- Comfortable working with large image and video datasets.
Experience with one or more of:
- Experience developing software for consumer products on Mobile SoCs, especially within the Android NDK framework.
- Few-shot Learning, Transfer Learning
Please note: This is a full-time position that will be remote initially (due to COVID-19) and based in either our NYC HQ or Santa Clara once safe to re-open the office.Apply for this job