Applied Scientist II, Demand Science

The Amazon Devices & Services Demand Science (DSci) team is seeking a scientist with strong and AI/ML and communication skills to help with demand forecasting and supply optimization for the entire Amazon device family of products, services, and accessories. Amazon Devices represents a highly complex space with 100+ products across several product categories (e-readers [Kindle], tablets [Fire Tablets], TV [Fire TV and remote], smart speakers and audio assistants [Echo], wifi routers [eero], and video doorbells and cameras [Ring and Blink]), for sales both online and in offline retailers globally.

We develop scalable and robust state-of-the-art ML, AI, and automation solutions - including transformer-based forecasting architectures, large language model (LLM)-powered agents, and agentic AI workflows - that learn from diverse data sources and power advanced predictive models. With better forecasts we drive down supply chain costs, enabling the offer of lower prices and better in-stock selection for our customers.

In this role, you will own the full lifecycle - from research through production deployment. You will drive end-to-end solutions: understanding business requirements, exploring large-scale historical data and ML models, building prototypes, developing conceptually new approaches (including AI-native experimentation and agent-driven automation), and partnering with engineering teams to deploy and maintain solutions in production. You will collaborate closely with scientists, engineering peers, and business stakeholders. You will be at the heart of a growing and exciting focus area for Amazon Devices and Services where AI is transforming how we forecast, validate, and optimize decisions.

You are an individual with strong science abilities, excellent communication skills, solid coding skills, and comfort with modern AI tools and frameworks. You will be responsible for researching, prototyping, experimenting, analyzing predictive models, implementing production-ready solutions, and developing AI-driven automation that progressively reduces manual intervention and enables hands-off-the-wheel science operations.

Key job responsibilities

• Build forecasting models from prototype through production, working closely with engineering to deploy at scale
• Find and integrate new data sources to improve forecast accuracy and coverage
• Design and deliver production-ready solutions for business-critical forecasting and optimization problems
• Define and track performance metrics - both technical (error rates, bias, coverage) and business (plan attainment, financial impact, reduction in manual overrides)
• Write and maintain clear technical documentation; present findings and recommendations to scientists, engineers, and business leaders
• Set team standards for methodology, code quality, experimentation rigor, and AI-assisted workflows

About the team

We are a focused science team looking for help with enhancing the demand forecasting framework we"re building and figuring out how to best solve the needs of both our internal stakeholders and cross-functional partners.