Staff Machine Learning Engineer, Ads Candidate Generation

Within the Ads candidate generation team, our mission is to bridge the gap between the aspirations of Pinners and the products offered by our partners. In this role, you will spearhead the developing and executing a visionary strategy to evolve our machine learning technology stack for Candidate Generators. You'll tackle cutting-edge challenges, including managing an ever-growing corpus of billions of shopping ads, exploring model architectures beyond the traditional two-tower structure, modeling users’ long-short interests, and implementing sequential transducers for generative recommendations. Additionally, you'll leverage large language models (LLMs) to enhance our recommendation systems and optimize dynamic quota allocation. Your work will be instrumental in advancing the ML models that form the backbone of our ads retrieval and delivery processes, effectively connecting Pinners with our partners' offerings in this unique marketplace.

What you’ll do:

• Be responsible for the development of state-of-the-art applied machine learning projects for ads candidate generation models 
• GPU-based ads retrieval system that extends beyond the traditional two-towers model architecture.
• Develop a unified retrieval model that leverages both organic and ad-related labels, as well as attributed conversion labels, to create a scenario-adaptive retrieval model.
• Collaborate with the team to explore opportunities for leveraging LLMs to enhance recommendation quality.

What we’re looking for:

• MS or PhD degree in Computer Science, Statistics or related field
• 6+ years of industry experience building large scale production recommendation or search systems
• 2+ years of experience leading projects/teams
• Strong mathematical skills with knowledge of statistical methods
• Cross-functional collaborator and strong communicator
• Experience in computational advertising is highly desirable, though not required
• Experience in advanced retrieval modeling is highly desirable, though not required
• Expertise in GPU model performance profiling and optimization is highly desirable, though not required
• Experience in utilizing large language models within production recommendation systems is highly desirable, though not required

This position is not eligible for relocation assistance.

#LI-HYBRID

#LI-SM4