Avaya Inc.

Patrick Tendick

Wen-Hua Ju

Joann J Ordille

Qian Yang

Shmuel Shaffer

David Skiba

George William Erhart

Lorraine Denby

Sheldon Joseph Davis

John Hamilton Slye

Reinhard P Klemm

Nina Tandon

Rodney Allen Thomson

David Weiss

Henry R Paddock

Roger I Krimstock

James M Landwehr

Fei Chen

David Zanoni

Rouba Ibrahim

Nilay Noyan

Matias Potel Feola

A device, system, and method use predictive analytics based on an application-centric approach. The method includes receiving a request from an interactive application interacting with a user utilizing a user device for action data indicating an action to be taken by the interactive application during a session with the user device, the request generated by the interactive application based on a decision point associated with the interactive application. The method includes receiving from the interactive application session data associated with the session and the user device. The method includes determining scoring package data associated with the request based on the session data, the scoring package data comprising a predictive model indicative of a plurality of actions to be performed by the interactive application. The method includes generating the action data based on the scoring package data and transmitting the action data to the interactive application.

Predictive analytics and services

Systems are provided to manage pauses, including on-hold, conditions in communications, such as text-based communications. Users may be provided with messages at a pace that, among other things, allows a particular cadence of messages to be maintained, even if a reply has been created. Accordingly, resources of a contact center may be better allocated to other matters requiring attention without causing undue customer dissatisfaction or the burden of handling additional messages. While voice-based 'hold' conditions may be managed with messages or music on hold, text-based communications may be managed with social messaging content and/or comfort messages.

Social messaging hold and wait treatment systems

In order to provide better service with self-service applications, metrics for a communication session between a user and a self-service application are identified. Based on the metrics, a pattern is determined that will likely predict an initial outcome of the communication session with the self-service application. In response to determining that the pattern will likely predict the initial outcome of the communication session with the self-service application, the way the user is managed in the communication session is changed. For example, if a particular pattern of IVR responses is detected that will likely lead to a caller abandoning a voice call, the voice call can be automatically transferred from the IVR system to a contact center agent. This provides for increased customer satisfaction and better utilization of contact center resources.

Prediction of contact center interactions

Text-based automated customer service agents provide an attractive contact center interaction for may customer-agent communications. However, the 'say anything' nature of text may result in additional problems not encountered in prompted interactions. A single misunderstanding or error may cause an automated communication to fail, however, by providing micro-tasks to a human agent, an automated communication may be maintained longer with benefit of a human agent's input. A lack of progress towards completion of a work item may cause the communication to be routed to a human agent and/or a different channel (e.g., voice, email, etc.). Accordingly, a micro-task or an opt-out operation may be provided to maintain the communication with an automated agent for as long as possible but yet provide improved customer service when necessary.

System initiated dialog adjustment

To provide better metrics for a contact center, the system determines when a contact center agent has been assigned to a plurality of agent skills in a skill group. A skill group is a set of skills for which the contact center seeks to calculate overall metrics across those skills. Information is received that the contact center agent is in a first state for the plurality of agent skills. For example, the contact center agent is available to support the different products. A status is calculated for the contact center agent in the skill group. The status for the contact center agent in the skill group is calculated based a minimum or maximum of the first state of the contact center agent in the skill group. The status of the first agent in the skill group is sent to a contact center administrator to better manage the contact center.

Management of contact center group metrics

Contact centers comprise agents, human and automated, interacting with customers over a network to accomplish an objective for the interaction. While a blunt, factual exchange may appear to be the most expeditious means to accomplish the objective, often this is not the case. Humans often respond better when the interaction comprises a mix of progress (e.g., portions of the interaction directed towards the objective) and banter (e.g., portions of the interaction not directed to the progress). Determining the mix may be provided by analyzing historic interactions and their success. A current interaction may be analyzed and when the mix is outside an acceptable range and an automated agent may be configured to alter the mix accordingly or, when the agent is a human, signal the agent to make the alteration. Success may be continually monitored and the target mix adjusted based on subsequent interactions.

On-topic monitor

A first communication session is created with a first application written in first language. A change is detected in first variables in the first application. The first variables are retrieved from the first application in response to detecting a change in the first variables. The first communication session is transferred to a second application written in a second language. A change is detected in a second variable in the second application. The second variable is retrieved from the second application in response to detecting the change in the second variable. A common pair of variables are matched between the first variables and the second variable. The first communication session and the transferred first communication session are associated as a common communication session in response to matching the common pair of variables. One or more of the first plurality of variables is then used by the second application.

Variable capture between applications

An agent rating prediction and routing mechanism provided by a contact center communication system for work assignment optimization is described along with various methods and mechanisms for administering the same. The prediction system proposed herein analyzes past agent performance, agent attributes, contact attributes, and customer attributes to calculate an outcome value and to provide a performance prediction for use in work item routing to contact center resources.

Agent rating prediction and routing

Aspects are directed towards methods including creating multiple states from at least two call center data sources; and combining at least two of the multiple states to obtain a data model, where the data model comprises an attribute related to the call center, and methods including collecting a set of call data from multiple sources, where the multiple sources comprise a management module source and at least one source not controlled by the management module, and analyzing the set of call data.

Reporting on call center data across multiple concurrent activity sources

A predictive model for abandoned calls provided by a contact center communication system for routing optimization is described along with various methods and mechanisms for administering the same. Based on multiple attributes, the system can calculate the probabilities, risks, and costs associated with abandonment and wait, taking into account the patience of a caller, region where the caller lives, agents available to handle the call, length of time that the caller has already been waiting, potential cost of having the caller and call type abandon, and potential total cost of having the caller wait to be served.

Predictive model for abandoned calls

Methods and apparatus are disclosed for notifying one or more users of information having a plurality of attributes. A plurality of subscribable objects are provided to one or more users, where each subscribable object comprises one or more attributes as required attributes and zero or more of the attributes as optional attributes. Subscription requests are received from one or more of the users based on at least one of the subscribable objects. The subscribable objects are based, for example, on a set of predefined templates. The predefined templates identify each of the plurality of attributes as a required attribute or an optional attribute. Each of the attributes is presented to a user using a display format selected based on a cardinality of the corresponding attribute.

Method and apparatus for a publish-subscribe system with templates for role-based view of subscriptions

A method is disclosed for associating a caller utterance with corresponding elements of a flowgraph that depicts a voice-response system. Thus, an analyst who selects an element in a flowgraph receives immediate access to those utterances&#x2014;and only those utterances&#x2014;that are associated with the selected element. The analyst can easily listen to the utterances and review statistics associated with them. The utterances are stored, retrieved, and presented in a systematic and organized way. In some embodiments in accordance with the present invention, an utterance is associated with the link between two events in the voice-response system. In some alternative embodiments, an utterance is associated with other data, including, but not limited to: the event that precedes it; the event that follows it; the telephone number of the caller.

Interactive flowgraph for analyzing caller utterances

Methods and apparatus are disclosed for automatic notification and response based on communication flow expressions having dynamic context. A message is provided from a sender to at least one recipient in accordance with a communication flow having a plurality of potential paths. The communication flow is controlled by a communication flow expression having a dynamic context, wherein the communication flow expression contains at least one primitive keyword indicating how the message should be processed. The dynamic context of the communication flow expression may be established using one or more labels or a context function. A number of primitive keywords are provided that allow a communication flow expression to react to a dynamic context.

Method and apparatus for automatic notification and response based on communication flow expressions having dynamic context

The present invention is directed to a data model for effecting transformations of entities from a first form of expression to a second different form of expression.

Database structures and administration techniques for generalized localization of database items

Methods and apparatus are disclosed for notifying one or more users of information, where the information has a plurality of attributes. A plurality of subscribable objects are provided to one or more of the users, where each of the subscribable objects comprises one or more of the plurality of attributes as required attributes and zero or more of the attributes as optional attributes. Subscription requests are received from one or more of the users based on at least one of the subscribable objects. The plurality of subscribable objects are based, for example, on a set of predefined templates. The predefined templates identify each of the plurality of attributes as a required attribute or an optional attribute. Each of the attributes is presented to a user using a display format selected based on a cardinality of the corresponding attribute. For example, the display format for a given attribute may comprise one or more of a user entry, a scrollable list and a check box.

Growing community of inventors

Basking Ridge, NJ, United States of America

Patrick Tendick