Applied Scientist, Minerva
Core
Build and deploy ML models for fraud, abuse, and content risk detection across Kindle Direct Publishing and Kindle Unlimited.
Role type
Senior Applied Scientist (Risk Science)
Builds
Production ML models for content risk and abuse prevention
Domain
E-commerce / Books / Trust & Safety
Deliverable
production ML models
Required skills
model building, Java, C++, Python, algorithms, data structures, numerical optimization, data mining, parallel/distributed computing, high-performance computing
Preferred skills
fraud detection, abuse prevention, model monitoring, data drift detection, graph-based ML, behavioral modeling, anomaly detection, LLMs, fine-tuning, embeddings, NLP, computer vision, multimodal models, large-scale data processing, patents/publications, generative AI tools
Technologies
Java, C++, Python, Spark, SQL
Responsibilities
Monitor production risk models and improve their quality; Design, train, and evaluate improved detection models; Build datasets, labels, and evaluation sets; Write clear readouts with recommendations for stakeholders