Datadog is now integrating the capabilities of the Google Vertex AI platform into its own application monitoring and security platform. This gives customers more insight, monitoring and analysis capabilities for the performance of their own AI and ML models.
The integration of Google Vertex AI into the Datadog platform extends the existing collaboration between the two parties, Datadog writes. This makes Datadog one of the first observability partners of the tech giant’s AI and ML platform.
The integration will soon allow AI teams and other developers to more easily monitor, analyze and further optimize the performance of their own custom AI and ML models.
Different monitoring features
More specifically, this will give developers more insight into predictive performance and resource utilization of their own AI and ML models. In a central dashboard, they discover more about “model prediction counts,” latency, errors and the resources used. For example, the memory, network, and compute power usage of these resources for the respective AI and ML models.
In addition, the integration of Google Vertex AI into the Datadog platform provides an opportunity to compare the performance of different AI models in production environments. Vertex AI users can also further identify data anomalies that may affect the reliability of their models.
Building on new features
Among other things, the integration builds on new functionality recently released by the Google Vertex AI platform. Datadog introduced its own generative AI observability functionality earlier this year. This functionality provides a combined service with the AI assistant Bits from Datadog, also the company’s mascot.
The integration between Google Vertex AI and the Datadog platform is available immediately.