This list contains only the countries for which job offers have been published in the selected language (e.g., in the French version, only job offers written in French are displayed, and in the English version, only those in English).
Universal Ads is looking for a Sr. Principal Machine Learning Engineer to lead the design, implementation, and evolution of our ML stack. You'll architect the systems that power our marketplace, including expected value models for performance advertising, pacing optimization, and real-time prediction services that determine how every impression is valued and delivered. This is a highly technical and strategic leadership role. You'll collaborate on the long-term vision for our machine learning roadmap, guide the development of robust and scalable infrastructure, and collaborate across product, data science, and engineering to ensure Universal Ads becomes the industry benchmark for efficient, intelligent ad delivery across CTV inventory
Job Responsibility:
Set vision and technical direction for the long-term ML strategy and roadmap for ad ranking at Universal Ads, including model architecture, infrastructure, and deployment frameworks
Architect and scale distributed ML systems capable of real-time decisioning across high-throughput ad environments
Partner with product and marketplace teams to align model performance with user and advertiser outcomes
Drive technical direction for cross-functional initiatives involving bidding algorithms, pacing, and system optimization
Shape technical culture, and ensure high standards for research rigor, reproducibility, and code quality
Establish strong experimentation and evaluation frameworks (A/B testing, counterfactual analysis, etc.) for model validation and pacing control
Represent Universal Ads externally in technical forums, publications, or conferences to shape the broader conversation around ML in advertising
Design and own ML data pipelines end-to-end, ensuring data cleanliness and freshness, and establishing the standards and infrastructure for reliable model inputs across ranking and pacing systems