Scientist / Senior Scientist, Ads Systems

Roblox Roblox · Consumer · San Mateo, CA · Data Science

Roblox is seeking a Scientist/Senior Scientist for their Ads Systems team. This role focuses on designing and leading offline model evaluation, contributing to ML feature and model improvements for ads, building evaluation and simulation models, and leading experimental design and analysis. The position is crucial for innovating advertising systems and operations within a new strategic area for Roblox.

What you'd actually do

  1. Design and lead offline model evaluation end-to-end (pipelines, frameworks, analysis).
  2. Contribute to ads ML feature and model improvements, collaborating with engineering to put these in production.
  3. Build offline evaluation and simulation models for our ads marketplace.
  4. Accelerate product development through your understanding of the underlying data and your ability to partner with product and technical leaders to provide insights.
  5. Lead experimental design and analysis for your ads domain.

Skills

Required

  • Domain experience in digital ads systems
  • ownership of critical analytic or production pipelines
  • 4+ years of industry or academic experience in data science, economics, analytics, or machine learning engineering
  • Expertise in data transformation and programming
  • proficiency in developing data pipelines through use of Spark, Hive, and Airflow
  • Experience with auctions and experimentation in market settings

Nice to have

  • Graduate education in a rigorous applied training including a graduate degree in economics, statistics, bioinformatics, cognitive science, physics, or similar

What the JD emphasized

  • critical analytic or production pipelines
  • ownership of critical analytic or production pipelines
  • design and lead offline model evaluation end-to-end
  • build offline evaluation and simulation models for our ads marketplace

Other signals

  • design and lead offline model evaluation end-to-end
  • contribute to ads ML feature and model improvements
  • build offline evaluation and simulation models for our ads marketplace
  • lead experimental design and analysis for your ads domain
  • design science roadmaps and experimentation agendas