Description

Join Amazon's Community Feedback organization and help transform how billions of customers make purchase decisions. As an Applied Scientist on our team, you'll build intelligent systems that understand, curate, and surface insights from billions of ratings and reviews. You'll drive innovation in multimodal ML, NLP, and agentic AI systems that power core Amazon experiences like review highlights, product insights, and Rufus, directly influencing unregretted purchase decisions for customers worldwide.

- Work with one of Amazon's largest and most diverse customer feedback corpora, applying advanced ML to real-world problems at unprecedented scale
- Collaborate with a senior, high-performing science team split between Berlin and Barcelona, with opportunities to mentor and grow
- Shape the future of customer trust through research in corpus understanding, content quality, and intelligent surfacing systems
- Partner with teams across Amazon (Search, Rufus, Personalization) to multiply your impact across the shopping journey

Key job responsibilities
- Design, develop, and deploy machine learning models for corpus understanding, including entity extraction, multimodal content analysis (text, images, videos), and structured insight generation at billion-scale
- Build and optimize ML systems for content quality assessment, moderation, ranking, and personalized surfacing that directly impact customer experience
- Drive end-to-end ownership of scientific initiatives from research and experimentation through production deployment and measurement
- Collaborate with product and engineering teams to translate complex ML solutions into scalable, reliable customer experiences
- Mentor junior scientists and engineers, providing technical guidance and fostering scientific excellence across the team

A day in the life
The Community Data & Science team focuses on understanding, shaping, structuring, presenting and vending Community Feedback to help customers use the wisdom of the community to make unregretted purchase decisions. We build and own ML solutions that help with i) shaping the community content corpus both in terms of quantity and quality, ii) converting the content into structured data and iii) presenting raw content and structured data to shoppers to inform unregretted purchase decisions. Today, our ML models support experiences like review solicitation and submission, content moderation, rating aggregation, review ranking and review summarization.