Have you ever ordered a product on Amazon and when that box with the smile arrived, you wondered how it got to you so fast? Have you wondered where it came from and what series of events helped make that quick delivery possible? If so, the Amazon Logistics (AMZL) Production Planning team is for you. We manage a supply chain of vehicles that help deliver billions of packages each year and will only be more vital in increasing the speed of delivery to our customers.
Are you interested in designing, engineering, implementing and owning the metrics model that drives continuous improvement for Amazon’s global delivery operations? AMZL delivery operations manage the delivery of tens of millions of products every week to Amazon’s customers and are on pace to become one of the largest parcel carriers in the world.
The AMZL PPT Analytics team seeks an enthusiastic, customer-obsessed Business Intelligence Engineer to help shape the next generation of LLM-driven analytics at Amazon. As a BIE on this team, you will define and evolve the foundation of how customers interact with data - building the data dictionary that powers the agent’s knowledge base, refining prompts and workflows, and driving self-service analytics experiences that scale across AMZL and R2L.
In this role, you will own the engineering behind core AI/BI initiatives that simplify how operators, planners, and leaders access insights. You will work on building scalable metadata layers, designing intelligent retrieval logic, and ensuring the agent’s responses are accurate, explainable, and aligned to business priorities.
To be successful, you need to be a proven engineer, communicator, and problem solver. You will deal with ambiguity, experiment with evolving LLM prompting strategies, and continuously raise the bar on how analytics are consumed. Great candidates are comfortable blending data, engineering, and generative AI techniques to deliver practical solutions—driving change not only through metrics but through the conversational interfaces and intelligent agents that customers use every day.
Key job responsibilities
Partner with technology, science, and analytics teams to design and evolve the UTR Analytics Agent, including data dictionary governance, knowledge base curation, and conversational workflows.
Engineer critical input/output metrics and align them with the agent’s prompt templates to ensure accurate, explainable, and scalable insights across the full operations spectrum.
Build and expand attribution use-cases within the Analytics Agent (e.g., DEA, VET/VTO, COH, QXG), enabling customers to diagnose drivers of operational misses, compare levers, and understand root causes through conversational interfaces.
Develop and automate process performance metrics that power the agent’s recommendations—enabling root-cause identification, trend detection, and prescriptive actions for global delivery stations.
Manage and prioritize multiple analytics agent features (metrics onboarding, attribution logic, prompt tuning) across concurrent projects and business lines.
Communicate findings, attribution insights, and agent adoption results with diverse internal stakeholders and leadership audiences, ensuring alignment and trust in agent-driven analytics.
About the team
The AMZL Production Planning Analytics team focuses on building automated solutions for our last mile network. We drive improved accuracy, reduced effort and reduced variation. We work with our Planning, Operations, Finance and People eXperience and Technology (PXT) teams to understand business requirements and deliver customized solutions to support day to day execution and the future vision of value add work for our partners.
