Melanie Li

Organizations deploying foundation model (FM)-powered applications face challenges in reducing hallucinations and reasoning errors. When FMs encounter ambiguous queries or multi-step instructions, they may guess, extrapolate, or conflate information, leading to unreliable outputs.

This workshop is intended for Amazon Connect users and data scientists and covers how to predict customer churn and proactively engage agents to retain customers. In this workshop, use real-time transcription and sentiment analysis from Contact Lens for Amazon Connect combined with customer profile data to train a machine learning model using Amazon SageMaker.