Abstract: A system and method of generating conversation topics using neural networks. The method includes providing, by a processing device, a plurality of conversations to a neural network to generate a plurality of clusters. The method includes selecting, for each cluster of the plurality of clusters, a topic and one or more keywords from one or more n-grams. The method includes evaluating, for each cluster of the plurality of clusters, the topic and the one or more keywords by searching historical conversations and current conversations to identify one or more conversations related to the cluster.

Abstract: Methods and systems for optimizing timing of sending communications to a user device are disclosed. The method includes identifying a first group of user objects having a first attribute. The method includes locating active checkpoints within a user path with each active checkpoint having a user object associated therewith. The method includes forming a second group of the user objects associated with the active checkpoints. The method includes removing, with a processing device, user objects from the first group and the second group that match a second attribute to form a third group with the remaining user objects form the first group and the second group. The method includes processing the active checkpoint for each user object of the third group in response to forming the third group.

Abstract: A system and method of managing links for tracking user interactions with content items. The method includes generating, by a processing device, a banner message comprising a content item, wherein the content item comprises one or more links, each link associated with a respective internet address. The method includes scanning the content item to identify at least one link of the one or more links. The method includes extracting the at least one link of the one or more links from the banner message. The method includes generating a tracking link based on the extracted link. The method includes generating an updated banner message based on the tracking link. The method includes sending the updated banner message to a client device to cause the client device to present the updated banner message on a display associated with the client device.

Abstract: Methods and systems for optimizing timing of sending communications to a user device are disclosed. The method includes identifying a first group of user objects having a first attribute. The method includes locating active checkpoints within a user path with each active checkpoint having a user object associated therewith. The method includes forming a second group of the user objects associated with the active checkpoints. The method includes removing, with a processing device, user objects from the first group and the second group that match a second attribute to form a third group with the remaining user objects form the first group and the second group. The method includes processing the active checkpoint for each user object of the third group in response to forming the third group.

Abstract: A system and a method of displaying a predictive answer in a conversation window, are described. The method includes transmitting, to a device executing an application, a first message to cause the device to display a conversation window within the application, wherein the conversation window has a character entry field. The method includes generating, by one or more processors, a predictive answer based on a Uniform Resource Locator (URL) associated with the conversation window or a URL shape of the URL, wherein the predictive answer includes a predicted inquiry of a user. The method includes transmitting, to the device, a second message to cause the device to display the predictive answer in the conversation window.

Abstract: A method and system includes identifying a user object at an initial node of a first user path. The method includes identifying a user object at an initial node of a first user path. The method includes determining whether to perform a company prioritization for the user object at the initial node. The method includes determining that the user object is linked to multiple companies responsive to determining whether to perform a company prioritization. The method includes selecting, by a processing device, a first company of the multiple companies instead of a second company of the multiple companies to prevent a duplicate message associated with the second company from being transmitted to a user device associated with the user object.

Abstract: A system and a method of displaying a predictive answer in a conversation window within a webpage, are described. The method includes displaying the conversation window and presenting the predictive answer within the conversation window prior to receiving an entry of a character in a character entry field of the conversation window. The predictive answer is determined based on a current context of a device displaying the webpage. The current context is at least partly defined by a Uniform Resources Locator (URL) of the webpage, and a URL shape of the URL is mapped to the predictive answer in a context model.

Abstract: A method and system includes identifying a user object at an initial node of a first user path. The method includes determining whether the first user path requires company prioritization responsive to identifying the user object at the initial node. The method includes determining whether the user object is linked to multiple companies responsive to determining that the first user path requires company prioritization. The method includes locking, by a processing device, a first company of the multiple companies to the user object in the first user path responsive to determining that the user object is linked to the multiple companies. The method includes verifying a data match for the first company of the user object at the initial node responsive to locking the first company to the user object. The method includes advancing the user object to a subsequent node of the first user path.