Efficient Anomaly Workflow Detection for End-to-end Provisioning Systems

A network system configured to train a machine learning (ML) model to provide anomaly detection simultaneously across elements along end-to-end provisioning flows, the network system comprising:

• at least one processor;
• and
• at least one non-transitory memory storing an application that, when executed by the at least one processor, causes the at least one processor to:
• receive a provisioning log captured from a plurality of provisioning flows over a predefined period of time, wherein the plurality of provisioning flows are based on service requests that are associated with a plurality of transaction types and a plurality of subscriber segments, wherein the provisioning log comprises time-stamped operations and time-stamped response codes associated with provisioning flow elements along the plurality of provisioning flows, and wherein the provisioning flow elements comprise the plurality of transaction types, the plurality of subscriber segments, a plurality of network provisioning catalogs (NPCs) including a translation of service plan features of the plurality of subscriber segments into provisioning instructions for a plurality of network elements (NEs), and the plurality of NEs,
• generate, based on the provisioning log, a training data set comprising, for each interval of a plurality of intervals across the predefined period of time, a plurality of success rates, each for a respective one of the provisioning flow elements in association with respective ones of the plurality of provisioning flows,
• train the ML model to approximate a probability density function of a distribution of the training data set, wherein the training comprises processing the training data set by the ML model, and
• provide the trained ML model for real-time provisioning anomaly detection in the network system.

At least one non-transitory computer-readable medium storing instructions that, when executed by at least one processor of a network system, cause the at least one processor to train a machine learning (ML) model to provide anomaly detection simultaneously across elements along end-to-end provisioning flows by:

• receiving a provisioning log captured from a plurality of provisioning flows over a predefined period of time, wherein the plurality of provisioning flows are based on service requests that are associated with a plurality of transaction types and a plurality of subscriber segments, wherein the provisioning log comprises time-stamped operations and time-stamped response codes associated with provisioning flow elements along the plurality of provisioning flows, and wherein the provisioning flow elements comprise the plurality of transaction types, the plurality of subscriber segments, a plurality of network provisioning catalogs (NPCs) including a translation of service plan features of the plurality of subscriber segments into provisioning instructions for a plurality of network elements (NEs), and the plurality of NEs;
• generating, based on the provisioning log, a training data set comprising, for each interval of a plurality of intervals across the predefined period of time, a plurality of success rates, each for a respective one of the provisioning flow elements in association with respective ones of the plurality of provisioning flows;
• training the ML model to approximate a probability density function of a distribution of the training data set, wherein the training comprises processing the training data set by the ML model;
• and
• providing the trained ML model for real-time provisioning anomaly detection in the network system.