ML Engineer - Ad Ranking & Deep Learning

Every day, tens of millions of people come to Roblox to explore, create, play, learn, and connect within 3D immersive digital experiences. Roblox's mission is to connect a billion people with optimism and civility across any device worldwide.

As a Staff Machine Learning Engineer on the Performance Advertising team
, you will develop deep learning-based models and large-scale recommendation systems to power Roblox's rapidly growing ad ranking platform. Working on an ad engine in its foundational stages, you will design world-class ML architectures from the ground up—building ad ranking algorithms, CTR/CVR prediction models, and personalized advertising experiences that deliver value to global advertisers and community members alike.

• Ad Ranking & Deep Learning: Drive the end-to-end design, implementation, and deployment of machine learning models for ad ranking algorithms, bid optimization, and ad retrieval.
• Large-Scale Recommendation Models: Architect and scale multi-stage recommendation models handling massive real-time inference workloads and sparse features.
• System Architecture & Specs: Author detailed technical specifications for new ML platform capabilities, offline feature pipelines, and real-time model serving infrastructure.
• Applied Research & Productionization: Balance cutting-edge ML research with practical, user-first product execution to integrate state-of-the-art model architectures into production.
• Cross-Functional Leadership: Partner across internal product, platform engineering, and data teams while serving as a technical ambassador for Roblox in the broader ML community.

• 5+ years of professional experience designing, building, and scaling machine learning solutions for real-world production environments.
• Bachelor of Science in Computer Science or a related technical discipline (Master’s degree or Ph.D. is a definite plus).

• Proven track record of shipping production-grade machine learning models for ad tech, search, recommendation systems, or personalized ranking.
• Strong technical leadership capabilities, with experience driving complex architectural decisions across multiple engineering organizations.
• User-centric engineering mindset with the ability to translate ambiguous product requirements into high-performing ML solutions.

• Experience with transformer-based model training, inference optimization, and product integration in ad-ranking or recommendation contexts.
• Deep understanding of deep learning frameworks (PyTorch, Tensor Flow), distributed training, vector search, and real-time serving pipelines.

• Workplace Model: Hybrid model requiring in-office attendance at San Mateo HQ on Tuesdays, Wednesdays, and Thursdays (Mondays/Fridays optional).
• Visa Sponsorship
  
  Note:
  
  Roblox currently may not be able to support future H-1B sponsorship or candidates holding certain U.S. visa categories for this specific role.