Multimodal ML Engineer - Health Sensing Evaluation
Cupertino, California, United States
Hardware
The Health Sensing team builds outstanding technologies to support our users in living their healthiest, happiest lives by providing them with objective, accurate, and timely information about their health and well-being. As part of the larger Sensor SW & Prototyping team, we take a multimodal approach, using a variety of sensors across HW platforms, such as camera, PPG, and natural languages.
Key Qualifications
- Motivation to ensure that powerful AI systems stay under human control for high stake usages such as Health
- Expertise within ML/DL fundamentals; experience in multimodal modeling is a plus
- Knowledge in generative machine learning techniques, for example, diffusion models, GANs, VAEs, and transformers
- Proficiency in Python and ML frameworks e.g. PyTorch, Tensorflow
- Write performant and clean code; familiar with software development standard methods/collaborations
- Independently running and analyzing ML experiments to diagnose problems and test changes aiming real improvements
- Excellent interpersonal skills; comfortable in a collaborative and ground breaking research environments
Description
In this role, you will be at the forefront of evaluating multimodal and generative models for real-world health/wellbeing applications on their objective quality and alignment with human intent and perception, such as truthfulness, adaptability, and model generalizability. You will work on data and evaluation pipeline of both human and synthetic data for model evaluation, leverage ML technologies such as reinforcement learning with human feedback and adversarial models.
Responsibilities:
Build the back-end system that generate and lead data from a variety of endpoints (e.g. health databases, human annotations, synthetic generations)
Build quality and eval pipeline, model experimentation such as adversarial testing
Build insights/interpretability tools; explore methods to understand and predict failure modes
Being a critical part of the core multimodal ML dev team, innovate solutions to enhance model performance on quality metrics such as robustness and generalizability
Team up with algorithm engineers to build end-to-end pipelines that prioritize rapid iterations in support for reliability of a complex multi-year project
Education & Experience
BS and a minimum of 3 years relevant industry experience