A method and apparatus for providing emergency assistance includes receiving audio, visual, or text data about an ongoing emergency at a public safety answering point, comparing that data to a database of other emergencies with an artificial intelligence engine, recognizing patterns in aggregated and correlated data by the artificial intelligence engine, and generating appropriate audio, visual, and text output for a human operator to respond to the emergency. An AI engine receives audio data, visual data, and text data related to ongoing emergencies being received at public safety answering points and compares it with data from other emergencies. The AI engine identifies key words and patterns and generates appropriate messages for operators. The system continuously generates feedback to the AI engine to aid in machine learning. The feedback includes: “the message was correct; continue drawing this conclusion”; and “the message was wrong; do not continue drawing this conclusion.”

Method for real-time automated call summarization comprises determining an issue of a caller based on at least one of a call transcript, an extracted intent from the call transcript, or a slot of the intent. Based on the issue, a resolution is determined, and further an action item to implement the resolution is determined. The determined resolution and the action item are displayed in a graphical user interface (GUI).

A method and system matching contact center agents and back office staff with a customer inquiry. Exemplary systems include an expert term extraction engine, a customer term extraction engine, and a matching engine to compare customer request terms to the expert terms from the customer term extraction engine. The comparison determines whether there is a match or potential match between the customer request terms and the stored expert terms. An exemplary system may also include a timer that communicates with one or more communication servers. Back office staff may assist contact center agents when one or more conditions are met, such as when a customer wait time exceeds a predetermined period or when there is no match or potential match between the customer request terms and the stored expert terms for contact center agents.

In the field of government engagement management, an agent guide or script-flow in an employee desktop web client is implemented. In such a system and method, when agents create interactions with clients they can follow a script-flow which will guide the agent through the interaction through a series of menu selections and automated sets of instructions. This feature of the government engagement management system allows existing customer investment from the rich desktop client or non-web client in developing specific scripts, that can also now function in the web client atmosphere. This system and method also enables an agent to handle calls with the web client more efficiently, and allows agents on the web client to automatically classify.

An enhanced information retrieval system takes a customer utterance and constructs a contextually-enriched content-based query allowing the system to retrieve the most relevant documents to assist an agent in a real-time conversation with the customer. Phrases in the utterance are classified as informational or non-informational using a machine learning system trained with phrases from prior conversations of multiple users. Content phrases are extracted from the informational phrases using keyword extraction (ranking noun phrases), intent/action extraction (semantic role labeling), and topic label extraction (clustering of historical logs). Emotional content is identified using a sequence tagging model and removed. Contextual information from prior conversations with this user is combined with the updated content phrases to create the contextually-enhanced content-based query, which can then be submitted to the information retrieval system.

A method and apparatus for providing an automated workflow guidance to an agent during an active call between the customer and an agent is disclosed. The method includes extracting, at a call analytics server (CAS), from a transcribed text of an audio of a call between a customer and an agent, a call context. The method further includes identifying, by the CAS, at least one workflow based on at least one of the call context, a call metadata, or a historical data, wherein the at least one workflow is identified from a plurality of workflows in a workflow repository remote to the CAS. The identified workflow is sent as guidance from the CAS to a graphical user interface (GUI) accessible by the agent, while the call is active.

A system of handling callers, uses a computer, that receives a call, assigns an operator to handle the call, and automatically recognizes at least one aspect of a voice from a caller, and automatically forms a response to be given to the caller. The caller is prevented from knowing they are speaking with a computer by receiving responses from multiple different similar sounding operators. The computer providing sound to the caller which interferes with the caller being able to determine that the caller has been placed on hold from the operator, e.g., an average of multiple people talking in the background. The computer also maintains a ledger of the call, where the ledger includes information about recognized voice from the caller, and responses which were given to the caller, for each of a plurality of exchanges which occur during the call and distributes that ledger.

Systems, methods, and software described herein provide enhancements for a voice communication service to forward communications to a secondary service. In one implementation, a method of operating a group communication service that facilitates voice communications for a group of end user devices includes exchanging voice communications between the group of end user devices, wherein the group of end user devices comprises a first end user device and at least one secondary end user device. The method further provides identifying, in a voice communication from the first end user device, a use of a key phrase and, in response to identifying the key phrase, forwarding at least a portion of the voice communication to a second service.

Agents, whether human agents or automated agents, may be provided with content to deliver to a customer during a communication. The content may have an emotional content, as well as a factual content, that may be appropriate or inappropriate for a particular communication with a customer. Agents may be prompted to provide the content and the emotional content but whether they do or not is not always certain. By determining a difference between actual emotional content and an expected emotional content and executing steps to correct such differences, communications that comprise emotional content outside of nominal range may be corrected, within the communication and/or in subsequent communications. Additionally, long-term trends for one or a plurality of agents may be identified and managed as appropriate.

Techniques and architectures for implementing a team of virtual assistants are described herein. The team may include multiple virtual assistants that are configured with different characteristics, such as different functionality, base language models, levels of training, visual appearances, personalities, and so on. The characteristics of the virtual assistants may be configured by trainers, end-users, and/or a virtual assistant service. The virtual assistants may be presented to end-users in conversation user interfaces to perform different tasks for the users in a conversational manner. The different virtual assistants may adapt to different contexts. The virtual assistants may additionally, or alternatively, interact with each other to carry out tasks for the users, which may be illustrated in conversation user interfaces.