Conversational knowledge graph powered virtual assistant for application performance management

A virtual assistant system for processing a natural language input using a conversational knowledge graph, the system including:

• a natural language understanding system that interfaces with a collaboration channel to receive natural language user query and translate the natural language query received from the user into corresponding intents;
• a fulfillment service system communicatively linked with the natural language understanding system to receive the intents from the natural language understanding system;
• and
• a conversational knowledge graph system communicatively linked with the fulfillment service system to provide conversational knowledge context information from a conversational knowledge graph to the fulfillment service system;
• wherein the fulfillment service system is configured to use the conversational knowledge context information to send customized back-end service calls to an application performance monitoring (APM) product configured to monitor performance of one or more applications, causing the APM product to execute the customized back-end service calls, receive a result of the APM product executing the customized back-end service calls, and provide a response to the natural language understanding system based on the received result;
• wherein the natural language understanding system is configured to translate the response from the fulfillment service system into a natural language response and provide the natural language translated response to the user through the collaboration channel;
• wherein the conversational knowledge graph comprises a plurality of conversations related to the APM product, each conversation of the conversational knowledge graph including a plurality of entities of interest to the user within a context of the respective conversation, and the plurality of conversations including, at least, a first conversation associated with metrics of the APM product;
• and
• wherein the fulfillment service system is further configured to:
• detect a change in a first metric of the APM product in the first conversation,
• in response to detecting the change in the first metric of the APM product, create an on-demand recommendation for the user based on the conversational knowledge context information received from the conversational knowledge graph, wherein the on-demand recommendation recommends an action to be taken by the user using the APM product, the action to be taken relating to one or more second metrics of the APM product in the first conversation that contribute to the first metric of the APM product, and
• include the created on-demand recommendation in the natural language translated response provided to the user.

A tangible, non-transitory, computer-readable medium storing program instructions that cause a computer to perform operations for processing a natural language input using a conversational knowledge graph, the operations including:

• receiving, at a natural language understanding system of a virtual assistant, a natural language query from a user;
• translating, by the natural language understanding system, the natural language query received from the user into corresponding intents;
• providing, by the natural language understanding system, the intents to a fulfillment service system of the virtual assistant;
• sending, by the fulfillment service system, a request to a conversational knowledge graph system of the virtual assistant, to retrieve conversational knowledge context information from a conversational knowledge graph;
• using, by the fulfillment service system, the retrieved conversational knowledge context information to customize back-end service calls to an application performance monitoring (APM) product configured to monitor performance of one or more applications;
• sending, by the fulfillment service system, the customized back-end service calls to the APM product, causing the APM product to execute the customized back-end service calls;
• receiving, by the fulfillment service system, a result of the APM product executing the customized back-end service calls;
• sending, by the fulfillment service system, the result of the APM product executing the customized back-end service calls in a response to the natural language understanding system;
• translating, by the natural language understanding system, the response from the fulfillment service system into a natural language response;
• and
• providing the natural language translated response to the user,
• wherein the conversational knowledge graph comprises a plurality of conversations related to the APM product, each conversation of the conversational knowledge graph including a plurality of entities of interest to the user within a context of the respective conversation, and the plurality of conversations including, at least, a first conversation associated with metrics of the APM product, and
• wherein the method further comprises:
• detecting, by the fulfillment service system, a change in a first metric of the APM product in the first conversation;
• in response to detecting the change in the first metric of the APM product, creating, by the fulfillment service system, an on-demand recommendation for the user based on the conversational knowledge context information received from the conversational knowledge graph, wherein the on-demand recommendation recommends an action to be taken by the user using the APM product, the action to be taken relating to one or more second metrics of the APM product in the first conversation that contribute to the first metric of the APM product;
• and
• including, by the fulfillment service system, the created on-demand recommendation in the natural language translated response provided to the user.