Principal Machine Learning Engineer, Ads & Promos Delivery
DoorDash · Sunnyvale, CA · 347 Ads Engineering
About this role
DoorDash is hiring a principal-level Machine Learning Engineer based in Sunnyvale, CA. The posting calls out experience with Python, Java, TensorFlow, PyTorch. Compensation is listed at $268,600–$395,000 per year.
- Role
- Machine Learning Engineer
- Function
- machine learning
- Level
- principal
- Track
- Tech leadership
- Employment
- Full-time
- Location
- Sunnyvale, CA
- Department
- 347 Ads Engineering
More roles at DoorDash
Job description
from DoorDash careers
About the Team
The Ads & Promos Delivery team powers the last-mile delivery of ads and promotions, two marketing products offered to merchants, connecting merchant intent with consumer demand across search and discovery experiences. As a Principal Engineer, you will lead the technical direction for AI-first experiences, including ranking and relevance systems that sit at the core of our ads marketplace and shape how ads are selected, ordered, and personalized in real time across all verticals.
You will design and build next-generation AI-first ranking systems using state-of-the-art techniques such as sequence modeling, deep learning, and large language models (LLMs). Your work will span query understanding, user and merchant representation learning, contextual relevance, and multi-objective optimization, balancing advertiser value, consumer experience, and marketplace health at scale.
You will set the long-term technical vision, drive cross-team alignment, and translate cutting-edge research into production systems that operate under strict latency, scale, and reliability constraints.
As DoorDash expands into 40+ global markets and new verticals such as Grocery and Retail, this role offers a rare opportunity to define how modern AI, including sequential models and LLM-powered decisioning, reshapes ads relevance in a closed-loop marketplace.
About the Role
- Apply state-of-the-art machine learning and LLM techniques to problems across personalization, query understanding, user and content understanding.