Enterprises are increasingly deploying AI solutions across various platforms, including AWS, Google Cloud Platform (GCP), and Microsoft Azure. AI agents generate operational traces across multiple hyperscalers and observability platforms, leading to fragmented monitoring data and lack of centralized observability and discovery capabilities.

Through this capability, the administrator can configure trace collection integrations with hyperscaler and cloud-native observability environments to offer a unified pipeline for enterprise AI trace ingestion. This enables AI steward to discovery and evaluate AI systems like AI Agents operating within the enterprise ecosystem by monitoring performance metrics, observability and security quality indicators, and other KPIs which are generated by ServiceNow observability service.

1. Multi-Cloud Trace Collection - Seamlessly collect AI agent trace data from AWS Cloudwatch, Azure Foundry and App Insights and Google Cloud Platform (GCP) through a single unified interface without the complexity of SDKs. These traces enable discovery of AI agents from these 3rd party systems as well.
2. Guided Trace Configuration - Set up trace collection in minutes using a guided workflow that walks you through various steps for each cloud hyperscaler. ServiceNow MID Servers provide secure, authenticated access to your cloud platforms while keeping sensitive credentials protected.
3. Automated Metric Generation - Your collected traces are transported to observability service which generates  actionable evaluation and security metrics if you have the evaluations and security capabilities enabled. View real-time insights into AI agent performance, quality, cost, and compliance directly in AI Control Tower, empowering you to make data-driven governance decisions.