A complete application in a minute
Tailored resume and cover letter, ready to send.
Amazon is seeking a Principal Product Manager - Technical for Alexa Personality & EQ to own the multi-year vision of Alexa's character, including voice, tone, and emotional intelligence. You will translate subjective questions into concrete strategy, ship model prompts, and iteratively improve Alexa’s interactions across devices and contexts.
You will work with frontier models, collaborate with design, science, and engineering to tune and evaluate models, and define measurable standards for
Alexa is Amazon's AI assistant, and customers have been talking to it for years. They already have a relationship with Alexa and an opinion about who Alexa is. The way Alexa talks, reacts, and connects is central to what makes it distinctive, and we are looking for a Principal Product Manager - Technical to own the vision for that: who Alexa is, how it engages, and how it makes millions of customers feel understood in every interaction. You will decide what that character becomes next. The work you own shapes every customer conversation, on every device, in every context.
Alexa is a distinct character, and that character is a core part of what makes Alexa valuable to customers. As the Principal PMT for Alexa Personality & EQ, you will define and protect it. You will own the multi-year vision for who Alexa is and how it expresses that identity across every customer interaction: its voice, tone, emotional intelligence, and humor. You will hold two goals in tension, making Alexa feel deeply personalized to each customer while remaining unmistakably and consistently itself.
This is a hands-in-the-product role, not a set-strategy-and-hand-off role. You'll author instructions and style prompts that produce Alexa's personality, ship those changes, and iterate week over week. The work demands real technical depth: you will work directly with frontier models, partner with design on how the character is expressed in the experience, and partner closely with science and engineering on how the models are tuned and evaluated, operating at the edge of what they can do today.
You will translate a subjective, high-judgment question, "who should Alexa be, and how should it feel to interact with Alexa," into a concrete and decomposed strategy that can be applied across Alexa, an explicit and measurable definition of Alexa's emotional intelligence, and shipped experiences. Building the mechanisms to measure that standard is as much of the job as defining it, through to the go/no-go calls on shipping model and prompt changes. These approaches are established and still evolving quickly, which means part of the work is inventing what comes next. Someone expecting pure creative strategy will be surprised by how much rigor this demands.