Machine Learning used by a firm to detect AWS cost anomalies
FittedCloud Inc. announced today new capabilities for anomaly detection in its offerings. These capabilities optimize costs for various Amazon Web Services Inc. products.
The firm stated that its tools now use sophisticated machine learning algorithms to detect when cloud resource provisioning of Amazon Elastic Block Store, Amazon Elastic Compute Cloud (Amazon EC2), and DynamoDB services departs from normal boundaries and alert customers.
These cost optimization anomalies could be caused by programming errors in automation scripts or exposed identity management (IAM), keys that could allow an attacker to create many public clouds compute and storage resources. This could lead to financial harm.
According to the company, its offerings now include new predictive analytics capabilities that can identify such cases.
The company provided specific examples. FittedCloud’s alerts for anomaly detection provide details such as the location and the credentials used to attack.
The company’s tools allow for near-real-time analysis and provide visual details of anomalies to alert customers.
According to the company, Prashant Parikh, a customer at Erwin, stated that FittedCloud was solving a significant problem — wasted AWS resources such compute and storage. Customers can save money by adapting the cloud to their needs using machine learning, which helps them avoid over-spending. AWS environments change quickly and create a lot of data. Customers can reduce the risk of being attacked with excessive resources by using anomaly detection. FittedCloud’s alerts make it easy to find the source of an anomaly and fix it.
Here is information about FittedCloud’s different service plans.