Abstract: A method of extending a tagging and masking of recognized entities in a conversation transcript to cover new entity types. The method is performed in relation to the conversation transcript in accordance with parameters defining a new entity type. The method may include the steps of: searching for one of the keyword lookup trigger or the entity lookup trigger and triggering a lookup process in response to a detection thereof; performing the triggered lookup process by searching the search range specified in the received parameters for the selected generic entity type and detecting the selected generic entity type therein; revising the conversation transcript so that the entity name of the detected selected generic entity type is replaced with the entity name for the new entity type; and outputting a revised version of the conversation transcript that includes the revision.

Abstract: A system for generating context-oriented summaries of audio conversations in real time includes a principal model with a transformer-based model and a natural language processing model and is pre-trained with training audio data to generate summaries. The pre-trained principal model receives real-time audio data, which is processed by the transformer-based model to generate a series of encoder hidden state tensors. The series of encoder hidden state tensors is processed by an independent self-attention layer to generate a new set of hidden states, which are processed by a mean pooling module to generate a single fixed-length tensor representation, which is processed by the decoder to generate context-oriented summaries of the real-time audio data.

Abstract: The present invention provides a method and system for segmenting an interaction. A computing device obtains an interaction data element that may include a plurality of lines, each representing a contribution of a specific participant to the interaction. Using at least one first language model, a list of line indices representing segmentation of the interaction data element into segments of distinct topics is generated. The list of line indices is evaluated to identify inconsistency of the segmentation. When inconsistency is identified, a refined segmentation is generated by at least one second language model, wherein the at least one second language model is more computationally intensive than the at least one first language model. The method and system enable efficient and accurate segmentation of interactions for improved analysis and searchability.

Abstract: Systems and methods for providing context-aware assistance in a contact center platform are provided. In particular, a computing system may determine user context data associated with a user of a contact center, the user context data including a current application context and receive a user input from the user on an interface view indicative of a requested operation, the requested operation including at least one of: retrieving information, managing one or more users, or managing work items.

Abstract: A system and method for identifying relevant intents from an inbound interaction of a conversation bot, to construct customized Interactive Voice Response (IVR) flows, is disclosed. The system includes a contextual embedding module, to generate a vector database and an input vector representation database and an utterance context identification module, to identify a list of potentially relevant utterances in the vector database which are contextually similar to the input vector representation database and re-rank the list of potentially relevant utterances to obtain a reranked list of potentially relevant utterances. The system further includes a pretrained LLM, to match a relevant intent to each of the potentially relevant utterances in the re-ranked list and assign intent valuation score, an IVR flow construction service, to construct customized IVR flows based on the intent with the highest intent valuation score and an IVR flow validation module, to validate the constructed customized IVR flow.

Abstract: A method for configurable caching, deduplication, and rate limit handling of user interface data requests according to an embodiment includes receiving a data request initiated by a user interface element of a user interface, determining whether a deduplication cache includes a deduplication entry associated with the data request, determining whether a main cache includes a cache entry associated with the data request in response to determining that the deduplication cache does not include the deduplication entry, determining whether the data request is associated with a rate-limited group of application programming interface endpoints in response to determining that the main cache does not include the cache entry or determining that the cache entry has expired, and making the data request to a backend system after a predefined server retry period associated with the rate-limited group has elapsed in response to determining that the data request is associated with the rate-limited group.

Abstract: Systems and methods for mining question-and-answer pairs from conversation data to update knowledge bases are provided. In particular, a computing system may obtain conversation data, extract a question-and-answer pair from the conversation data with meta data using a machine learning model, the question-and-answer pair including a question and an answer corresponding to the question, determine whether a knowledge base includes existing content similar to the question-and-answer pair, provide the question-and-answer pair on a graphical user interface with an indication of a presence of any existing content in the knowledge base that is similar to the question-and-answer pair, receive an input to integrate the question-and-answer pair with the knowledge base, and update the knowledge base to integrate the question-and-answer pair in accordance with the input.

Abstract: A method for operationally transitioning between a first and second model in a contact center. The method includes defining a transition window and a pre-transition window. The method includes, during the pre-transition window, using the first model to determine first outputs and, therefrom, determining first output probability distributions and using, in real-time, the first outputs operationally in relation to pre-transition customers.

Abstract: A method for incentivizing contact center agents with scheduling perks that includes: receiving a workload forecast and staffing plan for a current shift; receiving a service level target; receiving types of scheduling perks; performing a selection routine for determining a select scheduling perk for offering to a select agent; and sending a scheduling perk offer via electronic communication to the select agent. The selection routine may include: selecting a proposed scheduling perks and a proposed agent; modifying the staffing plan to create a proposed staffing plan that reflects the proposed scheduling perk; predicting a proposed forecasted target adherence using the modified staffing plan; determining whether the proposed forecasted target adherence satisfies a threshold defined by an acceptable forecasted target adherence and, if so, deeming the proposed scheduling perk as select scheduling perk and proposed agent as select agent.

Abstract: A method of integrating services of contact centers which utilize dissimilar call models according to an embodiment includes communicating resource data resource data indicative of a voice interaction with complex call routing and/or complex call transfer stored by the on-premises contact center to the cloud-based contact center, and synchronizing the resource data indicative of a voice interaction with complex call routing and/or complex call transfer between the on-premises contact center and the cloud-based contact center to maintain consistency between the resource data stored by the on-premises contact center and the resource data stored by the cloud-based contact center.

Abstract: A method for authenticating a contact center user to a contact center agent includes a cloud system, to which the agent is securely authenticated, triggering a sending of an electronic message to a user device. The user device has one or more biometric authentication features, which it requires to access a private key, of a key pair, securely stored on the user device. The electronic message contains a validation link or a trigger for a validation prompt. A passkey system stores an agent identifier, a user identifier, an interaction identifier, and a credential identifier in a database, in response receiving a positive authentication status from a business web server indicating that the business web server has authenticated the customer device by passkey. In response to the passkey system receiving the positive authentication status, the passkey system transmits an authentication status notification indicating the positive authentication status to the cloud system.

Abstract: A system for complexity assessments for agent performance analysis using a memory neural network according to an embodiment includes receiving a plurality of call recordings of calls between a contact center agent and one or more users, generating, for each call recording, a respective conversation summary of the call recording, analyzing each respective conversation summary using the memory neural network to determine a respective agent performance category for performance of the agent during the respective call, each respective agent performance category being selected from a plurality of predefined agent performance categories and associated with a respective call score, comparing each respective call score to a predefined threshold value and updating a topics array based on the respective conversation summary in response to determining that the respective call score is below the predefined threshold value, and determining a total performance score for the agent based on each respective call score.

Abstract: A method for generating an inverted search index for event sequence visualization may include determining, from an event sequence dataset containing event sequences and events, event pair combinations based on identified unique singleton events. The method may further include establishing a positional ordering for each singleton event and each event pair combination and generating a respective event sequence set for each event sequence. The method may also include determining whether each unique singleton event and each event pair combination exists or does not exist within each event sequence, and populating, for each respective event sequence set and according to the positional ordering, data indicative of whether each unique singleton event and each event pair combination exists or does not exist within the event sequence for the respective event sequence set. The method may also include generating the inverted search index including each respective event sequence set for each event sequence.

Abstract: A method for determining representative utterances from input utterances to generate paraphrased utterances with healthy intents which optimizes intent health in a bot flow architecture. The conversational bot flow architecture includes a machine learning model trained for natural language understanding (NLU) within a NLU domain that is defined by a collection of intents and sets of associated utterances.

Abstract: A method for error reduction in intent classification according to an embodiment includes determining a set of possible intent classes for a section of input text, mapping each of the possible intent classes into a respective group of a plurality of groups, determining which groups of the plurality of groups are contradictory to one or more other groups of the plurality of groups, splitting the section of input text into a plurality of sub-parts, performing intent classification on the section of input text to determine an intent of the section of input text and on each sub-part of the plurality of sub-parts to determine a respective intent of each sub-part, evaluating the determined intent of the section of input text and the respective intents of the sub-parts for intent contradiction, and overriding the determined intent of the section of input text in response to identifying an intent contradiction.

Abstract: The present invention relates generally to the technological field of search engines. More specifically, the present invention relates to detecting statistical irregularities in data and retrieving records pertaining to these irregularities. The invention represents a system and methods of retrieving records from an indexed database which provide an improvement of the technological field of search engines and information retrieval by making the detection and retrieval of records representing statistical irregularities in the database faster, more efficient, and less computationally complex.

Abstract: A method for performing time constrained queries for event sequence visualization may include receiving an event sequence dataset including event sequences. Each event sequence includes at least two sequentially occurring events. The method further includes generating a forward index having event pattern sets that include pattern occurrence data for each event sequence positioned according to a fixed positional order. The method includes generating a time constraint search structure, such as a B?tree, B+tree, or table, that includes keys and pointer values. The method also includes inserting, into the search structure, a sequence identifier corresponding to each event sequence based on the fixed positional order and time interval data for the at least two sequentially occurring events of each event sequence.