Staff Software Engineer, Machine Learning

About the role

Staff Machine Learning Engineer: Growth and Personalization

This role is central to DoorDash's retail expansion. The team is responsible for constructing the underlying intelligence that powers how customers discover and purchase goods across our network. You will architect the models that determine relevance for millions of shoppers. The work spans research, implementation, and long-term stewardship of production systems.

The position is based in a hybrid model. You will split time between the office and remote work. The office location is primarily San Francisco, California, with consideration for Sunnyvale, California. The compensation package reflects the market value for this role, with a base of $183,000 and an expected total of $271,500 for the 2024 plan year.

What you'll do

• Design and deliver machine learning features that directly influence the retail conversion funnel, moving from abstract hypothesis to stable, user-facing functionality.
• Own the full lifecycle of model development, including research, implementation, and long-term production stewardship.
• Design new ML modules that enhance search precision for retail categories such as grocery and convenience.
• Lead data-driven experiments to measure the impact of model changes across distinct retail segments.
• Mentor less experienced engineers and coordinate with product and design pods to ensure high-quality delivery.
• Design systems that adapt to changing shopper intent and a growing catalog, with a core focus on ranking and causal inference models.
• Build scalable training pipelines to support personalization, loyalty, and discovery initiatives.
• Work closely with product leaders to convert business objectives into testable hypotheses and success metrics.
• Ensure all solutions adhere to strict privacy and bias audit standards.
• Maintain production systems for deployment, logging, and performance monitoring, with reliability as a paramount concern.

Requirements

Candidates must bring significant industry experience, with eight or more years developing machine learning models that impact business metrics. You must have a track record of shipping solutions into production environments.

Proficiency with modern AI coding tools is essential. You will use tools such as Claude Code, Codex, and Cursor to design, generate, test, and release software.

A graduate degree is required. You must hold an M.S. or PhD in a quantitative discipline. Acceptable fields include Statistics, Computer Science, Mathematics, Operations Research, Physics, and Economics.

Expertise in causal inference and recommendation systems is mandatory. This includes both classical statistical methods and modern deep learning approaches. You are expected to write Python code regularly, with hands-on experience in PyTorch or TensorFlow in production required.

You must be able to explain complex technical concepts to non-technical stakeholders, using data and rigorous experimentation to validate ideas and guide larger strategic investments.

Practical Information

Please confirm all details on the official application page. The information provided here serves as a guide and may be updated based on feedback.

The role operates under a hybrid engagement model, allocating time between remote work and office presence. The primary office location is San Francisco, California, with consideration given to Sunnyvale, California for team proximity. This role is classified as Staff Software Engineer, Machine Learning and reports to the Machine Learning organization. The compensation structure for the 2024 plan year establishes a base salary of $183,000, with an expected total compensation of $271,500. Candidates must possess eight or more years of hands-on experience building and deploying machine learning models that directly influence business outcomes. You must demonstrate a proven ability to ship production-grade systems and maintain reliability in live environments.

You will leverage advanced tooling in the AI-assisted development space, utilizing platforms such as Claude Code, Codex, and Cursor throughout the software lifecycle. A graduate-level degree in a quantitative field is mandatory, with acceptable disciplines including but not limited to Statistics, Computer Science, Mathematics, Operations Research, Physics, and Economics. Mastery of causal inference methodologies and recommendation system architectures is a non-negotiable requirement. You will write Python on a daily basis and maintain hands-on production experience with PyTorch or TensorFlow. Communication skills are critical; you will regularly translate complex technical concepts for non-technical audiences and rely on empirical data and rigorous experimentation to validate hypotheses and drive strategic investment.

This position requires authorization to work in the United States, and sponsorship is not available for this role. The position is exempt from standard working hour policies, and the hybrid schedule requires physical attendance in the office for designated collaboration periods. The expected start date is aligned with business needs, and early engagement with the recruiting team is necessary to meet onboarding timelines.