Senior Applied Scientist, Agentic Ad AI

Amazon Science; Seattle (WA); On-site; USD 167,000 - 226,000; Full time

Job summary

Amazon Science is seeking an Applied Scientist III to set scientific direction for agentic AI applications that guide Amazon advertisers. You will own end‑to‑end agentic architecture, define research roadmaps with product and engineering leaders, and mentor scientists while delivering production‑grade code.

You will work on PM‑aligned, medium to large, ambiguous problems, building scalable systems across ad campaigns with a strong emphasis on safety, trust, and real‑world impact for advertisers

Qualifications

• 5+ years of building machine learning models for business applications.
• PhD or Master’s degree with 6+ years of applied research experience.
• Strong programming experience in Java, C++, Python or related languages.
• Experience with neural deep learning methods and ML.

Responsibilities

• Define the science vision for the agentic campaign management system and translate it into roadmaps.
• Build agentic systems that autonomously manage and optimize ad campaigns balancing ROI and shopper experience.
• Specify datasets and signals to train and evaluate agents including campaign and auction data.
• Write production‑quality code and build core components from prototype to launch.
• Own agentic architecture including planning, tool use, and multi‑agent orchestration with hands‑on coding.
• Define evaluation and safety methodology for agent workflows and drive reliability standards.
• Mentor scientists and engineers and represent the team in internal and external forums.

Skills

Machine learning experience

Leadership & mentorship

Education

PhD or Master’s + 6+ years applied research

Tools

Java

C++

Python

Distributed systems

Deep learning frameworks

Job description

Amazon Science is seeking an Applied Scientist III to set scientific direction for agentic AI applications that guide Amazon advertisers. You will own end‑to‑end agentic architecture, define research roadmaps with product and engineering leaders, and mentor scientists while delivering production‑grade code.

You will work on PM‑aligned, medium to large, ambiguous problems, building scalable systems across ad campaigns with a strong emphasis on safety, trust, and real‑world impact for advertisers