Overview In this role you will build agentic systems that optimize advertisers' campaigns across targeting, bids, budgets, and placements. You'll work with cross-functional teams to turn marketplace signals into intelligent, live-acting agents guiding campaign decisions. You'll push the design of planning, tool use, and long-horizon reasoning, delivering production-grade systems that operate at scale. This is a hands-on, research-driven role shaping the future of ad optimization at Amazon Ads.
Compensation / Benefits- health insurance (medical, dental, vision, prescription)
- RSUs and sign-on incentives
- 401(k) with company match
- paid time off and parental leave
- flexible spending accounts and mental health support
- career development opportunities
Responsibilities- Build agentic systems that optimize campaigns across all controls (targeting, bids, budgets, placements) and generalize across campaign types and advertiser goals
- Encode auction and marketplace dynamics to balance advertiser ROI, shopper experience, and marketplace health
- Define and curate datasets to train and evaluate agents, including campaign, marketplace, auction, and performance signals
- Develop core agent capabilities (planning, tool use with auction simulation, ML models, optimization routines) and production-quality code
- Establish evaluation and safety methods to enable trustworthy live-campaign action
- Collaborate with product and engineering partners and contribute to the scientific agenda of the team
- Share work with the broader community and contribute to a culture of innovation
Key requirements- 3+ years of building models for business applications
- PhD or Master's degree with 4+ years in CS, CE, ML or related field
- Proficiency in Java, C++, Python or related language
- Experience with algorithms and data structures, numerical optimization, data mining, and distributed computing
- collaboration with cross-functional teams
- strong problem-solving and analytical thinking
- clear technical communication
- agent design and reinforcement learning
- LLMs and tool integration
- auction simulations and optimization
