Explore a 19-minute conference talk on transforming fraud management in the insurance industry from reactive to preventive approaches. Discover how Manulife implemented a real-time, AI-driven system to combat fraud, which costs billions annually. Learn about the innovative solution and architecture that improved data extraction and anomaly detection using open-source vision, large language models, Azure Cognitive Search, Document AI, link analysis, and scaled traditional ML models within the Azure-Databricks platform. Gain insights into how this new system integrates data from multiple sources to provide risk scores for providers, members, and fraud rings, flagging potential fraudulent claims. Understand the impact of this transformation, including millions in avoided fraudulent claims, a 60+% boost in accuracy, and a 12X speedup. Delve into the journey, lessons learned, technical details, and future plans shared by Sabyasachi Mukherjee, AVP Advanced Analytics at Manulife. Access additional resources like the LLM Compact Guide and Big Book of MLOps to further enhance your knowledge in this field.
Overview
Syllabus
Reactive to Preventive: Managing Fraud with Databricks OpenCV and GenAI
Taught by
Databricks