Our Retrieval team builds shared machine learning capabilities that help Pinterest connect people with relevant ideas. We are looking for a Staff Machine Learning Engineer to advance how we develop those capabilities through agentic AI.

Your primary focus will be Reflex, our effort to automate and accelerate recommendation research and development. You will own a major technical area, building agents that investigate opportunities, implement changes, run experiments and learn from their results. You will work closely with other technical leaders while remaining hands-on with the hardest engineering and modeling problems.

You will also contribute to generative retrieval, with a particular focus on Semantic IDs: learning structured representations of content that recommendation models can generate and use to retrieve relevant items. This role combines applied AI research, production engineering and technical leadership, with opportunities to improve both our recommendation models and the way engineers build them.

What you'll do:

• Own and advance a major technical area in automated recommendation research within Reflex, collaborating with other technical leads and partner teams.
• Build LLM agents that connect hypothesis generation, implementation, experiment execution, evaluation and learning from previous results.
• Develop rigorous evaluations for agent behavior and research outputs, and use them to improve quality, reliability and efficiency.
• Contribute to generative retrieval models, particularly Semantic ID learning, evaluation and integration into production recommendation systems.
• Take promising ideas from research and prototypes through offline evaluation, online experimentation and production launches.
• Build reliable integrations with ML development infrastructure and make informed tradeoffs across model quality, iteration speed, compute cost and operational complexity.
• Raise technical quality through hands-on implementation, design and code reviews, documentation and mentorship.
• Work with research, recommendation and ML infrastructure teams to shape technical approaches, resolve dependencies and support adoption.
• Identify and apply relevant advances from research, and contribute new methods where existing approaches fall short.
  
  

What we're looking for:

• Bachelor's in Computer Science, Machine Learning or a related field, or equivalent practical experience.
• Minimum 8 years of experience building and shipping ML or AI systems, including substantial hands-on engineering and technical leadership.
• Demonstrated experience building LLM agents or other agentic AI systems for complex tasks, with a clear approach to evaluating and improving their performance.
• Strong ML fundamentals and depth in one or more relevant areas, such as applied LLMs, representation learning, recommendation systems, retrieval or ML systems.
• Experience taking research ideas into reliable production systems and evaluating their impact through meaningful experiments.
• Strong software engineering skills, including the ability to design maintainable systems, implement difficult components and conduct thorough technical reviews.
• Experience leading a substantial technical area, making decisions under ambiguity and collaborating effectively across teams.
• A track record of mentoring engineers and raising the quality of a team’s technical work.
• Clear written and verbal communication, including the ability to explain technical tradeoffs to different audiences.
• A disciplined approach to using AI in your own work, including verification of generated code, analysis and conclusions.

We'll teach you our stack and our products, and we're committed to your continued growth as a technical leader. This is an individual contributor role; prior people management isn't required.

Relocation Statement:

• This position is not eligible for relocation assistance. Visit our PinFlex page to learn more about our working model.

In-Office Requirement Statement:

• We let the type of work you do guide the collaboration style. That means we're not always working in an office, but we continue to gather for key moments of collaboration and connection.
• This role will need to be in the office for in-person collaboration 1-2 times/quarter and therefore can be situated anywhere in the country.

#LI-REMOTE

#LI-DM57