In June 2023, Google enhanced its Vertex AI platform, which provides AI and machine learning services, by incorporating generative AI support, allowing users to test and deploy large language models (LLMs). Datadog has announced an integration with Vertex AI, enabling users to monitor the health and performance of their LLM-powered services in production. This integration provides insights into key metrics such as network traffic, prediction errors, latency, and resource utilization through an easy-to-use dashboard. Users can track the standard RED metrics—rate, errors, and duration—to assess model performance and infrastructure health, and receive real-time alerts for anomalies in metrics like memory and CPU usage. The integration aims to offer a comprehensive view of AI deployments, helping users quickly identify and address issues, and ensuring efficient operation of generative AI-powered services.