Applied Scientist Ii, Foundation Model, Industrial Robotics Group

Amazon Amazon · Big Tech · Sunnyvale, CA · Research Science

Applied Scientist II role focused on developing foundation models for industrial robotics, integrating multi-modal learning, skill acquisition, perception, and environmental understanding. The role involves leveraging, adapting, and optimizing state-of-the-art models, conducting rigorous experimentation, building evaluation benchmarks, and contributing to data and training workflows. It requires strong programming skills in Python and experience in deep learning areas like computer vision, multimodal models, or RL for robotics.

What you'd actually do

  1. Leverage state-of-the-art models for targeted tasks, environments, and robot embodiments through fine-tuning and optimization.
  2. Execute rapid, rigorous experimentation with reproducible results and solid engineering practices, closing the gap between sim and real environments.
  3. Build and run capability evaluations/benchmarks to clearly profile performance, generalization, and failure modes.
  4. Contribute to the data and training workflow: collection/curation, dataset quality/provenance, and repeatable training recipes.
  5. Write clean, maintainable, well commented and documented code, contribute to training infrastructure, create tools for model evaluation and testing, and implement necessary APIs

Skills

Required

  • Python
  • deep learning
  • computer vision
  • multimodal models
  • imitation learning
  • RL for robotics
  • human-robot interaction
  • experiment design
  • results analysis
  • reproducible baselines
  • production-quality research code

Nice to have

  • C++
  • Java
  • Unix/Linux
  • professional software development

What the JD emphasized

  • PhD, or Master’s + 4+ years building ML models/algorithms in applied settings.
  • 2+ years hands-on experience in deep learning with strength in at least one: computer vision, multimodal models, imitation learning / RL for robotics, or human-robot interaction.

Other signals

  • foundation models
  • robotics
  • multi-modal learning
  • skill acquisition
  • perception
  • environmental understanding