Staff Machine Learning Engineer

Zigsaw

Staff Machine Learning Engineer at Pinterest leading ML strategy and developing recommendation systems for advertisers and sellers. Collaborating on AI-driven solutions in a dynamic team setting.

Posted 6/2/2026full-timeSan Francisco • California • 🇺🇸 United StatesLead💰 $189,308 - $389,753 per yearWebsite

Tech Stack

Tools & technologies

JavaPython

About the role

Key responsibilities & impact

Lead the design and implementation of large-scale recommendation and decisioning systems that power proactive advertiser and seller guidance across Ads Manager, Pinterest Business Assistant, Pinnacle, and sales productivity workflows.
Build ML foundations for a unified context layer and context agent that transforms campaign, account, performance, market, workflow, and interaction data into reusable signals for agentic experiences.
Own recommendation initiatives end-to-end, from problem framing, label and feedback design, feature pipelines, model development, and offline evaluation through production deployment, experimentation, and monitoring.
Develop evaluation and feedback loops that measure recommendation quality, user trust, action rates, business impact, and failure modes, then use those learnings to continuously improve models and agent behavior.
Apply modern ML techniques such as retrieval and ranking, embeddings, personalization, multi-objective optimization, contextual decisioning, and response modeling to business-critical advertiser and seller workflows.
Use AI to accelerate analysis, prototyping, documentation, and experimentation while applying strong judgment, testing, data validation, and review to ensure correctness, reliability, privacy, and customer trust.
Mentor engineers and raise the technical bar for ML development, experimentation rigor, responsible AI usage, and production-quality agentic systems across the organization.

Requirements

What you’ll need

7+ years of experience building and deploying large-scale ML systems in production (e.g., ads ranking, recommendation, Agentic AI, or search), with strong end-to-end ownership from problem scoping through evaluation and experimentation, and solid software engineering skills in at least one modern language (e.g., Python, Java) and large-scale data systems.
Degree in Computer Science, Mathematics, or a related technical field, or equivalent experience.
Strong end-to-end ML ownership, including problem scoping, data and label design, feature engineering, model training, production deployment, offline/online evaluation, experimentation, and monitoring.
Deep understanding of recommendation system architectures such as candidate generation, retrieval, ranking, re-ranking, embeddings, vector search, multi-task learning, calibration, contextual bandits, or reinforcement learning.
Proven Staff-level technical leadership as a hands-on IC, setting technical direction and driving multi-quarter ML and systems roadmaps, including aligning stakeholders on priorities, trade-offs, and execution plans.
Excellent cross-functional communication and collaboration skills, building strong partnerships with product, data science, infra, and partner ML teams to clarify ambiguous problem spaces, co-create solutions, and drive consensus with senior stakeholders.
Experience using AI coding assistants (e.g., Cursor, Claude Code) and LLM-powered productivity tools to accelerate development, experimentation, and data exploration, with a clear approach to validation, data protection, and critical review of AI-assisted work.

Benefits

Comp & perks

Information regarding the culture at Pinterest and benefits available for this position can be found here.

ATS Keywords

✓ Tailor your resume

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills & Tools

machine learningrecommendation systemsdata engineeringfeature engineeringmodel trainingproduction deploymentevaluationexperimentationPythonJava

Soft Skills

technical leadershipcross-functional communicationcollaborationproblem framingstakeholder alignmentmentoringjudgmentdata validationtrust buildingsolution co-creation