Drimva

Snap

Staff Research Scientist, User Modeling and Personalization

$195K–$343K

Bellevue, Washington · Staff · Full-time · Posted Sep 25, 2026

Experience
5+ years of industry or postdoctoral experience.
Education
PhD in computer science, machine learning, language technologies, or a related technical field such as statistics or mathematics, or equivalent years of experience.

About the job

Develop research and scalable machine-learning systems for user modeling and personalization, translating research into customized experiences across Snap.

What you'll do

  • Define and lead research in user modeling and personalization, including generative modeling, recommendation, information retrieval, and efficiency.
  • Build and evaluate scalable research prototypes and partner with engineering teams to apply research to real-world products.
  • Mentor teammates and interns and publish findings at leading conferences.

What they ask for

  • PhD in computer science, machine learning, language technologies, or a related technical field such as statistics or mathematics, or equivalent years of experience.
  • 5+ years of industry or postdoctoral experience.
  • Track record of publications in leading machine-learning, information-retrieval, or language venues.
  • Experience with distributed multi-node and multi-GPU ML training, inference, and experimentation, and applying language models to generative search, ranking, or personalization.

Nice to have

  • Experience with large-scale ML research or equivalent open-source work, and large-scale data processing using cloud solutions such as Google Cloud, AWS, or Azure.
  • Familiarity with post-training, preference optimization, search or recommendation interaction data, and recommender systems.

Pay

$229,000-$343,000 annually (base salary); $218,000-$326,000 annually (base salary); $195,000-$292,000 annually (base salary)

Tools

Python, PyTorch, Google Cloud, AWS, Azure

AI focus

Research and build generative and language models, recommender systems, information retrieval, and representation-learning approaches for user modeling and personalized experiences.

Model-written summary · the original posting is authoritative

First seen Oct 28, 2026 · Last verified Nov 5, 2026 (UTC)