🌟Quantum for Azure - Introducing the New Cost Anomalies Feature

We’re excited to share the first release of a new Cost Anomalies feature in Quantum, designed to help you quickly identify unusual changes in your cloud spend and take action early.

This is an initial version, and we’ll continue to enhance it over time. Here’s a quick guide to help you get started:

🔍 Getting Started: Where to Find Anomalies

Navigate to the Cost Anomalies menu option — your central hub for spotting unusual cost patterns.

undefined

You can sort by highest cost or highest severity to focus on the most impactful anomalies first.
Anomalies are currently detected at the Resource Group level, and include both cost increases and decreases (this will be refined in future updates).

🚨 Types of Anomalies Detected

The system uses several detection methods to flag potential issues:

Statistical Outlier – Spots sudden spikes in cost.
Cost Projection Risk – Warns when future costs are projected to be high (based on linear trend analysis).
Trend Shift – Highlights changes between the current and previous two-week periods.
Pattern Break – Detects deviations from the usual cost pattern.
Multiple Alerts – Indicates when more than one of the above has triggered.

🕵️‍♀️ Investigating an Anomaly

Click on any anomaly to explore it in more detail. You’ll find several helpful tabs:

✅ Overview

Alert Reasons – Understand why this was flagged.
Key Metrics – See average daily cost, total cost over the last 14 days, and projected daily cost for the next 14 days.
Anomaly Classification– you can label the anomaly as:
- Expected Change
- False Positive
- Not Important
- (Default: Unexpected Cost)

This helps improve future accuracy and lets you track anomalies more effectively.

📈 Cost Trends

View a graph of daily costs for the Resource Group.
This helps you visually assess whether the change is a one-off or part of a trend.
Note: This version uses simple linear trend analysis and may not yet account for regular weekly cost cycles.

🔮 Projection & Risk

Explore possible future cost scenarios.
Includes confidence scores and recommendations for addressing the anomaly.

⚙️ Detection Config

See the current sensitivity settings (how much costs needs to change to trigger an anomaly alert, and those filtered out).
Note: For this early release, the system is tuned to be more sensitive to help with fine-tuning.

🧭 Next Steps

If an anomaly looks concerning:

Go to the Cost Allocation section.
Drill down into the specific Resource Group to identify the services or resources driving the change.

📌 A Few Final Notes

This is an early release, and we’re still in an experimental phase.
Expect improvements and new capabilities over the coming months.
Documentation and training videos will be available in future versions.

If you have any questions or need help using the feature, just reach out to the Bytes Service Desk