Developing Gen AI applications with Amazon Bedrock has been made easier thanks to Temporal, which removes the friction and pain of hosting reliable, scalable models. However, building production-grade Gen AI applications still poses several challenges, including handling user inputs, managing long-running sessions, and providing a conversational experience. Temporal can be used to solve these problems by modeling each conversation as a workflow that is guaranteed to execute to completion, with features such as caching/persistence, queues, tracing, observability, and logging tools. By using Temporal's event history and queries, developers can easily view the execution history of a workflow, including the inputs and outputs of activities, making it invaluable for rapid debugging and investigation. With Temporal Workflows, users can model entire conversations as entities that may last forever, providing greater visibility into the history of workflows and durability guarantees to ensure workflow history and state is always maintained.