Introduction
AI applications in production generate large volumes of telemetry data across distributed services, but raw data alone doesn't provide actionable insight. This module guides you through analyzing application telemetry with Azure Monitor logs and metrics to detect failures, identify performance trends, and maintain operational visibility for AI solutions on Azure.
Imagine you're a developer building a document processing pipeline for an enterprise content moderation AI service. The system consists of four services: an ingestion API that receives document uploads, a classification service that categorizes content using a trained model, an extraction service that identifies key entities, and a moderation service that flags policy violations. After deploying to production, the team notices that some documents take over 30 seconds to process, but there's no way to determine which service causes the delay. Occasionally, the moderation service returns errors for specific document types, and the team only discovers these failures when users report them. Your client expects a dashboard that shows real-time pipeline health, with alerts that notify the on-call team within five minutes of a failure spike. The client also needs the ability to investigate incidents interactively, drilling into specific time windows and filtering by document type or service. Azure Monitor provides the query language, visualization tools, and alerting capabilities to meet all of these requirements.
After completing this module, you'll be able to:
- Write KQL queries to retrieve and analyze application telemetry from Application Insights.
- Explore log data to identify error patterns, performance bottlenecks, and trends in application behavior.
- Build Azure dashboards that display key telemetry metrics and log query results for operational monitoring.
- Create Azure Monitor Workbooks for interactive, parameter-driven telemetry analysis.
- Configure alert rules that detect application failures, performance degradation, and anomalies.
Note
All code examples in this module use KQL queries against Application Insights log tables. The Azure Monitor query experience is updated regularly, and the recommendation is to visit the Azure Monitor logs documentation for the most up-to-date information.
Note
We recognize that different people like to learn in different ways. You can choose to complete this module in video-based format or you can read the content as text and images. The text contains greater detail than the videos, so in some cases you might want to refer to it as supplemental material to the video presentation.