Senior ML Engineer, End-to-End Delivery AI

DoorDash USA; San Francisco, Sunnyvale, Seattle (CA, CA, WA); On-site; USD 137,000 - 299,000; Full time

Benefits offered by this job

401(k) matching

Paid parental leave

Medical, dental & vision

PTO & sick leave

Wellness benefits

Commuter benefits match

Job summary

DoorDash is seeking a Machine Learning Engineer for the Drive team to own ML systems end-to-end from feature engineering to deployment and monitoring. You will build models for delivery ETA, pickup ETA, and merchant prep-time estimation, and apply reinforcement learning and optimization to improve logistics decisions.

The role involves building AI-native experiences using DL, LLMs, and multimodal signals, plus rigorous online experimentation and cross-functional collaboration.

Qualifications

• 5+ years of industry experience building production ML systems.
• Production ML models using modern DL frameworks such as PyTorch.
• Strong software engineering with Python and ML infra tooling.
• Experience with estimation, ranking, prediction, optimization, or decision-making at scale.

Responsibilities

• Own ML systems end-to-end from feature engineering to deployment and monitoring.
• Build models for delivery ETA, pickup ETA, merchant prep-time estimation, and order release.
• Develop DL models using large-scale signals and apply RL/optimization techniques.
• Collaborate with engineers, PMs, data scientists, and platform teams.

Skills

Python

PyTorch

Deep Learning

Reinforcement Learning

Optimization / OR

LLMs / VLMs

Production ML systems

Delivery/logistics domain

Education

Bachelor/Master/PhD in CS/ML/Related field

Tools

Spark

Airflow

Job description

DoorDash is seeking a Machine Learning Engineer for the Drive team to own ML systems end-to-end from feature engineering to deployment and monitoring. You will build models for delivery ETA, pickup ETA, and merchant prep-time estimation, and apply reinforcement learning and optimization to improve logistics decisions.

The role involves building AI-native experiences using DL, LLMs, and multimodal signals, plus rigorous online experimentation and cross-functional collaboration.