Senior Data Scientist, Creator Platform

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

Senior Data Scientist on the Creator Platform team at Roblox, focusing on building and maintaining ML systems for creator reputation, anti-abuse, and rewards. The role involves driving creator data foundations, enabling agentic and AI-powered analytics, and defining reporting metrics. Requires strong ML, experimentation, causal inference, and data foundation expertise, with a focus on trust & safety and AI for analytics.

What you'd actually do

  1. Own Creator Reputation & Publish Gating: Serve as the primary DS owner for the Creator Reputation Model and Good Standing score — ML systems that estimate creator trustworthiness and gate publishing access for new or alt accounts before they accumulate creator history. Maintain and improve the auto-approval model for DevEx requests.
  2. Build & Maintain Anti-Abuse Models: Lead anti-abuse modeling for creator rewards and creator analytics, ensuring the integrity of creator incentive systems and identifying bad actors at scale.
  3. Drive Creator Data Foundations: Serve as DRI for core creator data infrastructure — documenting all key creator tables, designing trustworthy data contracts, and building the semantic layers and analytical frameworks that enable reliable self-serve analysis across the team.
  4. Enable Agentic & AI-Powered Analytics: Lead DS AI enablement for the creator org — building skills, evaluation frameworks, and scalable AI-enabled workflows that reduce analyst toil and elevate the team's analytical capacity.
  5. Own Creator Analytics Reporting: Define top-line success metrics and own measurement for creator analytics, providing leadership with a clear, consistent view of creator health and platform performance.

Skills

Required

  • Expert-level SQL
  • Expert-level Python
  • ML model development
  • ML model evaluation
  • ML model maintenance
  • Trust & Safety ML
  • Anti-abuse ML
  • Reputation systems ML
  • Experimentation design
  • Causal inference
  • Data foundation design
  • Semantic layer design
  • Data contracts
  • Agentic analytics
  • AI-enabled analytics
  • Evaluation frameworks for AI outputs
  • Communication skills
  • Influence skills
  • Problem framing
  • Data infrastructure

Nice to have

  • Experience with ecosystem or marketplace settings
  • Mentorship
  • Code review
  • Cross-functional partnership

What the JD emphasized

  • primary strategic data partner
  • owning end-to-end problem framing
  • ML models, especially for trust & safety, anti-abuse, or reputation systems
  • complex, mediated settings where simple A/B tests don't tell the full story
  • data foundation layer for AI
  • agentic and self-serve analytics use cases
  • enough engineering fluency to work effectively across data pipelines, model outputs, and production-facing systems
  • operating independently in zero-to-one environments

Other signals

  • ML systems that estimate creator trustworthiness and gate publishing access
  • Lead anti-abuse modeling for creator rewards and creator analytics
  • Enable Agentic & AI-Powered Analytics
  • building skills, evaluation frameworks, and scalable AI-enabled workflows