Staff Research Scientist, User Modeling and Personalization

Snap Snap · Consumer · Bellevue, WA +1

Research Scientist role focused on user modeling and personalization, involving generative models, recommender systems, and information retrieval. The role requires translating research into real-world ML applications, building scalable prototypes, and publishing findings. Experience with large-scale ML training and inference, and applying language models for generative search, ranking, and personalization is expected.

What you'd actually do

  1. Formulate and derive a research agenda in the user modeling and personalization domains, including generative modeling, recommendation systems, information retrieval, and efficiency
  2. Partner with engineering teams to translate research to business impact for real-world ML applications used by millions of Snapchatters
  3. Build scalable research prototypes and evaluate them in large-scale machine learning scenarios
  4. Share your expertise with teammates and interns
  5. Publish your findings at top conferences

Skills

Required

  • PhD in computer science, machine learning, language technologies or related technical field such as statistics, mathematics, or equivalent years of experience
  • 5+ years of industry or postdoctoral experience
  • Strong technical knowledge of machine learning, information retrieval, personalization, language and state-of-the-art deep learning literature
  • Demonstrated ability in defining, leading and executing challenging research projects
  • Strong computer science fundamentals, problem-solving and engineering skills (Python, PyTorch)
  • Pragmatic, hands-on approach to research with a drive to build working prototypes rather than solely rely on theoretical exploration
  • Proven ability to mentor interns, students and junior researchers
  • Experience with distributed (multi-node and multi-GPU) ML model training, inference and experimentation
  • Experience applying language models in the context of generative search, ranking and/or personalization

Nice to have

  • Experience with large-scale machine learning problems in an academic or industrial research lab, or equivalent open-source experience
  • Experience with large-scale data processing, collection or synthesis using machine learning frameworks on Enterprise Cloud solutions like Google Cloud, AWS, and/or Azure
  • Familiarity with post-training, preference optimization, working with large-scale search or recommendation interaction data, and recommender systems
  • Demonstrated ability to transform cutting-edge research into tangible product improvements

What the JD emphasized

  • Track record of publications in top machine learning, information retrieval or language venues
  • Experience applying language models in the context of generative search, ranking and/or personalization

Other signals

  • User Modeling
  • Personalization
  • Generative Models
  • Recommender Systems
  • Information Retrieval
  • Large-scale ML