Sr. Staff Machine Learning Engineer, Agentic Ads
Lead technical strategy for ads ranking and bidding models at Pinterest using large-scale ML and agentic development loops. Requires 8+ years applied ML experience with ranking/ads systems, deep expertise in recommendation techniques, and proven ability to influence cross-functional partners.
What you’ll do
- Lead the technical strategy for advancing ads ranking and bidding models using large scale ML and agentic development loops.
- Design, build and launch production models and signals that improve ads quality, relevance and long term value.
- Build and refine training, evaluation and feedback pipelines that continuously learn from online behavior, experimentation results and advertiser outcomes.
- Use AI to accelerate analysis and iteration while applying judgment and verification to ensure correctness and quality.
- Own the technical strategy for advancing ads ranking models, with a focus on AI/agentic development loops and automation.
- Build and refine agentic training and evaluation pipelines that continuously learn from feedback signals, offline and production behavior.
- Partner closely with product, applied science, and infra teams to translate business objectives into modeling roadmaps and success metrics.
- Ensure models meet the bar for reliability, safety, fairness, and compliance with privacy and policy constraints.
- Drive technical design reviews, set modeling best practices, and make high-impact architectural decisions across the ads ML stack.
- Mentor and uplevel other MLEs and scientists in advanced modeling techniques, tooling, and agentic workflows.
- Shape product outcomes at massive scale, with direct impact on advertiser value, Pinner experience, and company revenue.
- Partner with senior leaders across product, research, and infra, gaining broad visibility and influence on ads strategy.
- Grow as a technical leader by setting modeling standards, mentoring senior talent, and driving multi-quarter roadmaps across teams.
- Access Pinterest’s rich multimodal data and modern ML/infra stack to experiment quickly and ship ambitious ideas.
What we’re looking for
- Minimum of 8 years of experience in applied machine learning including large scale ranking or ads systems.
- Deep expertise in modern recommendation and ranking techniques such as gradient boosted trees and deep learning based ranking models.
- Experience building and operating ML systems at scale in languages such as Python or C++ and with modern data and experimentation platforms.
- Demonstrated ability to use AI to improve speed and quality in your day to day workflow for modeling, experimentation and analysis.
- Experience leading cross functional initiatives and influencing senior partners across product, engineering and research.
- Bachelor’s/Master’s degree in a relevant field such as computer science or statistics, or equivalent experience.
- High integrity and ownership in how you handle data, use AI responsibly and remain accountable for final decisions and deliverables.
Staff Machine Learning Engineer
Staff ML Engineer leading end-to-end identity verification ML systems including document authenticity, face matching, liveness detection, GNN-based identity graphs, and behavioral risk models. Requires 8+ years production ML experience and domain expertise in biometrics or fraud detection.
Staff ML Engineer
Founding Staff ML Engineer building production ML systems for governance, security, and agentic platform capabilities at Docker. Owns architecture, data pipelines, evaluation, and model lifecycle while mentoring the growing team.
Principal Engineer, AI Platform
Principal Engineer setting technical vision and building AI/ML infrastructure for Generative AI and Recommender Systems at Pinterest, scaling to hundreds of millions of inferences per second. Requires deep expertise in distributed systems and proven cross-org technical leadership.
Senior Research Engineer, Post-training & Evaluation
Own evaluation science and post-training methodology for Reddit's foundational LLMs. Define benchmarks, design model-as-a-judge systems, and set SFT recipes that turn base models into safe, Reddit-native endpoints.