Staff Machine Learning Engineer, Personalization
SpotifyJob Description
📋 Description Own and improve the ML models powering Home feed and Shortcuts. Design, build, and ship personalized recommendations for millions of listeners. Build content recommendation systems for agentic and AI-powered experiences. Train, fine-tune, evaluate large language models using SFT, distillation, LoRA. Partner with product managers, engineers, data scientists, and designers to run experiments. Drive A/B testing, monitoring, evaluation, and optimization for quality, reliability, and cost. 🎯 Requirements 8+ years building and deploying ML systems in production. Deep expertise in recommendation systems, ranking, personalization, or large-scale discovery. Strong Python proficiency and hands-on PyTorch experience. LLM training, fine-tuning, evaluation, incl. SFT, distillation, LoRA. Experience with large-scale inference: latency, reliability, cost optimization. Distributed ML workloads with Ray, FSDP, HSDP, or similar. Data pipelines and orchestration with Flyte, Airflow, BigQuery, cloud storage. 🎁 Benefits Flexible remote work within North America; EST-friendly collaboration. Health insurance, parental leave, 401(k), meal allowance, PTO and holidays. Equal opportunity employer with accommodations available.