Lead content strategy for agentic products. Define and own how Spotify communicates through LLMs by setting voice, personality, and tone standards. Build content frameworks used by composition and engineering.
Design, author, and evaluate prompts and system instructions for generative features. Build reusable prompt templates, experiments, and guardrails with product, engineering, and safety teams.
Write and iterate on high-quality, in-product language (microcopy, help text, error states, onboarding, notifications, tooltips, naming) that is accessible, inclusive, and measurable.
Partner with product managers, designers, researchers, and engineers to translate user and business needs into content requirements; participate in discovery, design critiques, and localization planning.
Partner with engineers to design and run human-in-the-loop evaluations. You’ll define what "good" looks like for AI-generated text and use those insights to tune model performance.
Requirements
You’re an experienced content designer and UX writer, with a portfolio that shows you can solve complex problems through language.
You have a background in conversation design and character development, and bringing brands to life through applied voice and tone guidelines.
You have hands-on experience working with LLMs, prompt design, and evaluation. You understand the difference between a zero-shot and few-shot prompt, and you know how to iterate on instructions to get a specific tonal output and avoid hallucinations.
You can navigate ambiguity, collaborate across teams and subject areas, and tackle complex flows that span states, surfaces, or systems.
Bonus: Experience with LLM-as-a-Judge frameworks, evaluation scaling, or basic scripting for testing prompts in notebooks & native or near-native proficiency in a second language.