Artificial intelligence news.
Straight from the original announcements.

Datadog expands Azure monitoring for AI workloads and cloud costs

November 17, 2026By the AI News Bites Editorial TeamSource: Datadog
Datadog expands Azure monitoring for AI workloads and cloud costs
Image: Datadog

Datadog is expanding its Azure monitoring coverage with 55 additional integrations for Azure Monitor services, alongside tools for tracking AI workloads and cloud spending. The company says the updates are designed to help engineers investigate performance issues and identify the sources of rising Azure and AI costs.

Key points
  • Datadog is adding 55 integrations for Azure Monitor services.
  • Azure OpenAI dashboards and monitors cover rate limits, capacity, failures, latency and token volumes.
  • Microsoft Foundry spending is available in Datadog’s AI Costs view.
  • Azure commitments can be tracked alongside broader Azure spending.
  • Advanced Blob Storage metrics show usage and blob counts by container.

Monitoring Azure AI applications across the stack

Datadog says its expanded capabilities cover Azure AI workloads from model requests and agent execution to the data services and infrastructure behind them. This is intended to help engineers distinguish between possible causes of slowdowns, including Azure OpenAI rate limits, repeated agent tool calls and slow database queries.

The Azure Cognitive Services integration includes an out-of-the-box dashboard and recommended monitors for Azure OpenAI deployments. Engineers can monitor azure.cognitiveservices_accounts.blocked_calls to detect requests rejected because of rate limits. Recommended monitors can also alert when provisioned deployments exceed 100% capacity, when more than 10% of calls fail, or when latency rises.

The Azure OpenAI dashboard includes prompt and response token volumes, which Datadog says can help teams identify workloads contributing to model usage and inference costs.

Tracing agents, databases and data dependencies

Datadog Agent Observability traces conversations, LLM calls, tool executions and retrieval steps. Engineers can examine individual sessions alongside error rate, p95 duration, estimated cost and token usage. The Summary view can also highlight issues such as repeated errors from a tool.

The company also describes monitoring for services that support agents. Azure AI Search coverage includes a recommended monitor for elevated search latency, while integrations for Cosmos DB for MongoDB vCore and Azure Redis Cache include monitors that fire when storage or memory exceeds 90%. Datadog Database Monitoring provides query-level visibility into PostgreSQL and MySQL workloads.

Agent Console, available in Preview, tracks usage, cost and latency across coding agents such as GitHub Copilot and Datadog Bits AI agents. Datadog GPU Monitoring covers GPU utilization and idle capacity across Azure, other clouds and on-premises environments.

Tools for controlling Azure and AI spending

Datadog Cloud Cost Management brings Microsoft Foundry spending into its AI Costs view alongside providers such as Anthropic and OpenAI. Teams can analyze spending by provider, model and user, then filter by provider, model and resource group to investigate individual workloads.

The service also consolidates Azure commitment programs, including Savings Plans, Reserved Instances and Databricks prepurchases. When spending changes, users can open Cost Change Summary to identify associated teams and tags, ask Bits to investigate the increase, and create a notebook, monitor or report from the analysis.

Datadog says its generally available Azure Data Explorer destination for Observability Pipelines can send logs directly to Azure Data Explorer without an extra hop through Azure Event Hubs. Teams can route detections to Microsoft Sentinel, long-term analytics to Azure Data Explorer, and operational data to Datadog for real-time troubleshooting and alerting.

More detailed Blob Storage visibility

Datadog is also using Azure Monitor advanced platform metrics to show usage and blob counts for individual Azure Blob Storage containers. The company says this addresses a limitation in account-level reporting, which could show that costs were rising without identifying the responsible container, application or team.

Engineers can rank containers by size and recent growth, account for previous blob versions and soft-deleted objects that continue consuming capacity, and monitor azure.storage.container_blob_capacity by container. They can also use a forecast monitor to receive an alert before a container crosses a threshold.

In Datadog’s example, the app-logs container adds close to 900 GiB per hour while other containers remain comparatively flat, and its blob count rises past 10 million. The company says overlaying deployment events can help determine whether such a jump coincided with an application change.

Related news

NVIDIA and Microsoft Extend AI Infrastructure From Azure to Windows PCsTop story
Chips and infrastructureNovember 19, 2026, 21:00

NVIDIA and Microsoft Extend AI Infrastructure From Azure to Windows PCs

NVIDIA and Microsoft are expanding their accelerated computing partnership across Azure, industrial systems and Windows PCs, adding access to Vera Rubin infrastructure…

Datadog certifies MinIO AIStor for on-premises BYOC observability
BusinessNovember 18, 2026

Datadog certifies MinIO AIStor for on-premises BYOC observability

Datadog has certified MinIO AIStor as an on-premises foundation for Datadog Bring Your Own Cloud, giving organizations a way to store telemetry on infrastructure they…

Microsoft Introduces Azure Catalyst for Agent-Led Modernization
Products and toolsNovember 17, 2026, 19:35

Microsoft Introduces Azure Catalyst for Agent-Led Modernization

Microsoft is introducing Azure Catalyst, an agent-led solution designed to modernize applications, data and infrastructure, alongside new Azure capabilities for building…