Spotify's Personalization team aims to make content discovery effortless and enjoyable for listeners by understanding music, podcasts, and audiobooks to deliver exceptional recommendations. The Search team specifically focuses on improving how listeners find and rediscover content, driving exploration and discovery, and connecting fans with creators through natural language understanding, personalization, and generative AI at scale.
Requirements
- hands-on experience training and operating transformer models in production, or strong interest in doing so
- production experience developing large-scale ML systems in Java, Scala, Python, or similar languages
- Experience with PyTorch or TensorFlow is a strong plus
- You have a strong background in machine learning and recommender systems, bridging research and user impact
Responsibilities
- Define and drive the ML technical strategy for Search, focusing on retrieval, ranking, and generative approaches
- Build models that improve query understanding, personalization, and relevance across Spotify’s Search experiences
- Prototype and productionize new modeling approaches at scale, serving hundreds of millions of users worldwide
- Lead high-impact projects from ideation through deployment, setting best practices for ML development, testing, evaluation, and experimentation
- Partner with tech leaders and stakeholders to influence priorities and ensure long-term scalability and impact
- Stay engaged with the broader ML and Search research community, applying emerging trends to Spotify’s challenges
- Build the next generation of Spotify’s recommendation systems, user representations, and supporting technical architecture
Other
- Collaborate with a cross-functional agile team spanning user research, design, data science, product management, and engineering
- You are comfortable navigating ambiguity and leading high-impact projects from start to finish
- You’re a systems thinker and strong communicator who can align and influence technical and product stakeholders
- You care deeply about agile processes, data-driven development, and reliability
- You’re eager to apply emerging ML trends, particularly in LLMs and generative recommenders, to Spotify’s challenges
