Sarah Aamiri

Led a team of Data Scientists in developing an AI-driven solution for insurers - Directed a team of 5 Data Scientists to build and maintain a document processing pipeline for health and travel insurers, leveraging OCR, document classification, and information extraction - Drove continuous improvements to the Claims Document Decisions solution, enhancing scalability and reducing configuration time from 6 months to 2 months - Transitioned project-focused approach to a product-oriented roadmap, ensuring long-term alignment with client needs and business objectives - Led client meetings to understand technical requirements and drive data science solutions that improved decision-making on complex insurance claims

Contributed to the industrialization of the Claims Decision solution for major clients - Built and industrialized the document processing pipeline (OCR, classification, information extraction, and reimbursement rules) for the first large-scale client - Developed KPIs to measure and track solution performance in production, ensuring real-time monitoring - Helped transition the solution from a proof of concept (PoC) to a production-ready product, streamlining deployment and scaling efforts

Developed data science models and tools to automate insurance decision processes (for Claims Decisions and Fraud Detection) Claims Decisions solution : - Built a real-time damage estimator using web scraping, enabling dynamic claims estimation in home insurance - Designed and implemented ETL processes to feed the decision engine with structured data, optimizing data flows for decision-making Fraud Detection solution : - Developed an image duplicates detection tool using Convolutional Neural Networks (CNNs), with invariance to document transformations (rotation, cropping, image-quality) improving document fraud detection