Monitor your AWS generative AI Stack with Datadog


As organizations increasingly leverage generative AI in their applications, ensuring end-to-end observability throughout the development and deployment lifecycle becomes crucial. This webinar showcases how to achieve comprehensive observability when deploying generative AI applications on AWS using Amazon Bedrock and Datadog.

Amazon Bedrock streamlines the development and scaling of large language models (LLMs), enabling teams to experiment with foundation models, customize them with their data, and securely integrate them into applications. Complementing this, Datadog’s solution provides deep insights into the behavior, performance, and cost of LLM-powered applications.

Through discussion and a live demo, attendees will learn how to monitor advanced LLM applications using Amazon Bedrock, while leveraging Datadog to monitor and troubleshoot real-world issues such as unexpected responses, model cost spikes, and performance degradations. The webinar highlights best practices for maintaining reliability, positive end-user experiences, and responsible AI practices throughout the application lifecycle.

By combining the power of Amazon Bedrock and Datadog, organizations can confidently deploy and scale generative AI applications while ensuring end-to-end observability, enabling them to unlock the full potential of this transformative technology.

In this webinar, you will learn:

- How Datadog can help solve for complexities created by generative AI and LLM’s

- How the Datadog SageMaker integration can access resource metrics—including CPU, GPU, memory, and network usage data—for all training and inference nodes.

- Learn about Datadog’s 2 new OOTB dashboards to better monitor model endpoints and jobs, where customers receive immediate time to value by gaining insights into resource utilization, error, and latency metrics.