Senior ML Engineer - AI Agents (Hybrid: Remote/On-site)

DoorDash High Volume; San Francisco (CA); Hybride; USD 168 000 - 247 000; Plein temps

Avantages offerts par ce poste

401(k) with employer matching

16 weeks paid parental leave

Wellness benefits

Commuter benefits match

PTO and sick leave

Résumé du poste

DoorDash is seeking a Senior Machine Learning Engineer for the Agentic Foundations team, focusing on vertical agent development. You’ll own the lifecycle of flagship agents like Ask Assistant, optimizing quality, latency, and cost through eval-driven loops and targeted prompts and tools.

You will partner with teams on agent memory, product representations, and post-training of small language models, translating research into production agents that benefit millions of customers.

Qualifications

• 3+ years of industry ML experience with hands-on work shipping LLM-based agents and eval-driven improvements.
• Experience with post-training or fine-tuning of open-weight models (SFT, RL) and small language models.
• Strong foundation in NLP and ML; proficient in Python and frameworks such as PyTorch or TensorFlow.
• MS or PhD in Computer Science, Statistics, Math, or equivalent practical experience.

Responsabilités

• Raise the quality of consumer agents by building and refining eval harnesses and prompts.
• Balance quality, latency, and cost by conducting rigorous experiments including fine‑tuning and deploying SLMs.
• Launch new agent experiences for Merchants and Dashers from exploration to production.
• Turn foundations into product impact by advancing agent memory and product representations for steerable generative recommendations.
• Shape the ML-driven roadmap for our grocery and retail delivery business in collaboration with leaders.

Connaissances

LLM-based agents

Prompting & context management

Eval-driven experimentation

Python expertise

Formation

MS or PhD in CS/Stats/Math

Outils

PyTorch

TensorFlow

Description du poste

DoorDash is seeking a Senior Machine Learning Engineer for the Agentic Foundations team, focusing on vertical agent development. You’ll own the lifecycle of flagship agents like Ask Assistant, optimizing quality, latency, and cost through eval-driven loops and targeted prompts and tools.

You will partner with teams on agent memory, product representations, and post-training of small language models, translating research into production agents that benefit millions of customers.