Vantage Enhances Anomaly Detection with Resource Attribution

Today, Vantage is launching an enhancement for Anomaly Detection that automatically associates anomalies with the specific resource that drove them. Customers can now identify not only when a Cost Anomaly occurs, but also precisely which resource is responsible, streamlining the process of addressing unexpected cloud spend. This feature is available to all customers beginning today, with no additional configuration required.

Cost Anomalies are now filtered for the resource that drove the anomaly.

Anomaly Detection on Vantage is powered by a machine learning model that looks for deviations to the upside at a service-category level (e.g., there was a significant change to Data Transfer for CloudFront). When an anomaly is detected, customers could open a Cost Report and identify the day the anomaly occurred and the service-category it occurred in. While this information was valuable, it often required additional investigations to determine the specific resource causing the anomaly, leading to delays in remediation and potential additional cost overruns.

Now, with the introduction of Resource Attribution for Anomaly Detection, Vantage automatically identifies and displays the specific resource driving a Cost Anomaly. When viewing a Cost Anomaly with an associated resource, the Cost Report will be filtered for the resource that was detected as causing the anomaly and highlight the corresponding day in the Cost Report’s bar chart, making it easy identify and reduces the time to resolution. Resource information will also be passed in any Anomaly Alerts, such as notifications through Slack or Jira for attention.

This functionality is now available to all customers beginning today. To see the new anomaly experience, head to any Cost Report and click the “Anomalies” tab at the top of the report. For more information on anomaly detection in Vantage, see the product documentation.

Frequently Asked Questions

1. What is being launched today?

Today, Vantage is launching the ability to tie anomalies to specific resources that caused them. Upon detecting a Cost Anomaly, Vantage will now automatically index resources that drove the increase in cost to identify if a specific resource is responsible for the change in cost.

2. Who is the customer?

Any Vantage customer that uses Anomaly Detection today can take advantage of this enhancement.

3. How much does this cost?

There is no additional cost to tie anomalies to specific resources that caused them.

4. I currently use anomalies in Vantage, what do I need to do to enable this?

Nothing: this is now the default experience for anomalies which are triggered by a specific resource.

5. What providers support this new functionality?

Any provider that provides resource level granularity can utilize this feature. At time of publishing, the following Vantage providers supported are:

• AWS
• Azure
• GCP
• Oracle Cloud
• Kubernetes
• Datadog
• Snowflake
• Databricks
• MongoDB
• Clickhouse
• Temporal
• Github
• Linode
• Confluent
• PlanetScale
• Fastly

6. When I click an anomaly from the anomaly detection list on a Cost Report, what is the default view?

When a resource can be attributed to an anomaly, you will be shown the Cost Report filtered for the Service, Category, and Resource the anomaly was detected for, with the day of the anomaly highlighted.

If the anomaly cannot be tied to a specific resource, the existing behavior will remain where you will be shown the Cost Report filtered to the service and category the anomaly occurred for, with the day of the anomaly highlighted. You can perform additional filters and group-by’s to identify root causes for these anomalies.

7. For anomalies tied to a resource, can I still group by other dimensions, just as I could before?

Yes, you can change filter and group by dimensions to do further investigation into any anomaly.

8. I have anomalies alerts set up via Slack, Microsoft Teams, Jira, or email, what will change with my experience?

You will not need to make any changes to your existing anomaly alerts. For anomalies that have a resource detected, this information will now be included in your alert.

9. Will resources continue to be associated to the anomaly when it is archived?

Yes, you will continue to be able to view anomalies attributed to resources after they are archived.

10. Will the API endpoint serving anomaly alerts include the resources that triggered the alert?

Yes, any resource information will be passed in the resources parameter as a part of the API response for anomalies that have resource attribution. See the below example

{
      "token": "anmly_alrt_1234567890",
      "created_at": "2025-05-13T06:27:43Z",
      "alerted_at": null,
      "category": "Compute Instance",
      "service": "AmazonEC2",
      "provider": "aws",
      "amount": "592.9343999949",
      "previous_amount": "521.1768974475",
      "seven_day_average": "498.6210299847",
      "status": "archived",
      "feedback": null,
      "resources": [
        "i-123456789abcd"
      ],
      "cost_report_token": "rprt_0987654321"
    }

If there is no resource attributed to the anomaly, the resources parameter will be blank.

11. Will previously triggered anomalies be retroactively associated with resources?

Yes, Vantage will re-process the previous 6 months of anomalies to add resources where applicable.

12. Where can I find more information or get help?

You can read more in our Anomaly Detection Documentation, or reach out to support@vantage.sh if you require any troubleshooting.