Machine Learning based anomaly detection in network traffic

Company Profile

Leading networking company that specializes in software-defined networking, providing an end-to-end solution that both simplifies and secures the WAN/branch office network.

Business Situation

The client was looking to create a way to add more long term value to their clients by leveraging technology like ML.  By working with GCP, they wanted to help improve certain areas of its core business by offering additional services to its customer base.  This helps improve the accuracy and speed of solving network problems by pinpointing, prioritizing, and offering prescriptive solutions to network issues.

Google Cloud Implementation

SpringML team was tasked to build an anomaly detection model for the client. A supervised machine learning model was built to predict network alarms. The client provided detailed logs that track network activity.  Such predictions benefit client’s customers as they can take action proactively. SpringML worked on an initial phase to prove the model. The SpringML team worked closely with the client’s engineers so that they can improve and operationalize the model.