Applied Scientist, Amazon Music - Catalog Quality
Core
Design and develop end-to-end systems to detect, measure, and remediate quality issues in music metadata (artist info, track attributes, versions, content tags) using Generative AI, classical ML, NLP, and Computer Vision.
Role type
Applied Scientist (Catalog Quality)
Builds
Scalable AI models, data pipelines, and model-serving systems for music catalog metadata quality.
Domain
Music streaming, metadata quality, Generative AI, Classical ML, NLP, Computer Vision.
Deliverable
production ML models
Required skills
Machine learning, deep learning, LLMs, Agentic AI, algorithms and data structures, numerical optimization, data mining, parallel and distributed computing, high-performance computing, predictive learning, statistical approaches, scalable data pipelines, model evaluation.
Preferred skills
Unix/Linux, professional software development.
Technologies
Generative AI, classical ML, Natural Language Processing, Computer Vision, LLMs, Agentic AI, Java, C++, Python.
Responsibilities
Collaborate with cross-functional teams to frame business problems as ML tasks; create scalable solutions using ML/deep learning/LLMs; analyze large datasets to automate/optimize processes; design and evaluate AI models for predictive learning; implement novel ML/statistical approaches; build scalable data pipelines and model-serving systems; analyze experimental results to refine models.
Seniority
Mid-Senior, hands-on IC