Abstract: Systems and methods of the present disclosure enable database search. The systems and/or methods may include receiving a search query that includes an input document having text. Word embeddings are generated within the input document, where the word embeddings include vector representations of words in the text of the input document. An average input document word embedding vector is determined for the word embeddings of the input document. A set of stored documents is accessed, where each stored document includes a stored text has a particular average stored document word embedding vector. A similarity model is used to determine a similarity metric measuring the similarity between the input document and each stored document based on the average input document word embedding vector and the particular average stored document word embedding vector of each stored document.

Abstract: Methods and systems disclosed herein describe customizing searching. Search queries may be customized according to a user's preferences. A user may emphasize or indicate that additional weight should be given to one or more terms in a search query. Terms that are weighted higher may have a larger impact on the results that are returned in response to the search query. In addition to changing the terms in a search query, a user may provide a weight for each term. Each term in a search query may be weighted to varying degrees, giving a user more control over the results that are returned. The weights may be used with machine learning techniques to generate a vector representation of a search query. The vector representation of the search query may be compared with vector representations of search objects to determine results that match the search query.

Abstract: Systems and computer-implemented methods disclosed herein relate to detecting errors in manually entered data. In one embodiment, the system can identify a named entity automatically from a conversation between a customer and service agent with a named entity recognition model that employs natural language processing and machine learning to detect a word or string of words in the conversation that corresponds to a named entity category. In another embodiment, the system can determine whether data entered into a field on a service platform by the service agent includes an error by comparing the data entered with the named entity. In another embodiment, the system can transmit an alert to the service agent through the service platform when there is a mismatch between the named entity and the data entered.

Abstract: In some embodiments, the present disclosure provides an exemplary method that may include steps of monitoring, a conversation script between a call center agent and a customer; utilizing, a speech-to-text deep machine learning model to transcribe the audio call to text; utilizing, a natural language processing deep machine learning model to map intent mappings of the audio call; utilizing, a similarity measurement model to determine a semantic similarity between predefined intent mappings and the intent mappings of the call audio text; determining, an error based on the semantic similarity in the intent mapping call audio text; determining a training session based on the error.

Abstract: Disclosed embodiments may include a method for automatically providing customized financial card incentives where the system can identify if a transaction is eligible for additional bonus rewards by determining if the user is a student using a clustering algorithm with user similarity data to generate a probability. Then, if the probability is above a predetermined threshold, a machine learning model with comprehensive user data classifies the user as a student or non-student. Once the user is determined to be a student, the transaction is verified as being an educational purchase using the transaction data or by having the user provide an image of a receipt. Once the transaction is verified, the transaction or items within the transaction qualify for additional rewards that are applied to the user's account. Users who are not identified as students may qualify for a standard amount of rewards.

Abstract: A system and method for allowing a single live customer service agent to simultaneously serve multiple customers. According to various embodiments, a virtual agent operates at the front end to receive and attempt to handle customer issues. The virtual agent employs speech recognition and intent mapping in order to generate a proposed response that attempts to identify and resolve customer issues. In some scenarios, the proposed response includes both a response message and a response action to be taken. A chat history and the proposed response is then provided to the live agent. The live agent reviews the information provided, and determines whether the proposed response is appropriate. The live agent then approves the response or revises or replaces the response. The final response is sent back to the virtual agent for processing and providing to the customer.

Abstract: Systems and methods of the present disclosure enable database search. The systems and/or methods may include receiving a search query that includes an input document having text. Word embeddings are generated within the input document, where the word embeddings include vector representations of words in the text of the input document. An average input document word embedding vector is determined for the word embeddings of the input document. A set of stored documents is accessed, where each stored document includes a stored text has a particular average stored document word embedding vector. A similarity model is used to determine a similarity metric measuring the similarity between the input document and each stored document based on the average input document word embedding vector and the particular average stored document word embedding vector of each stored document.

Abstract: Systems, methods, and non-transitory computer-readable media can be configured to perform operations comprising determining capabilities associated with a plurality of devices in a communication session, determining one or more media channels for transmission of media between the plurality of devices based on the capabilities, and facilitating the transmission of media between the plurality of devices via the one or more media channels.

Abstract: A processor may receive data and generate a quantified representation of the data by processing the data using at least one machine learning (ML) algorithm, the quantified representation of the data indicating a sentiment of content of the data. The processor may automatically revise the content of the communications data. The revising may include determining a reaction to the content of the communications data, generating a quantified representation of the reaction, determining a difference between the quantified representation of the reaction and the quantified representation of the communications data, identifying, based on the difference, a portion of the content having an unintended sentiment, and replacing the portion of the content with different content.

Abstract: In some implementations, a system may capture audio from a call between a calling device and a called device. The system may filter the captured audio to generate a background audio layer. The system may generate an audio footprint that is a representation of sound in the background audio layer. The system may determine that the audio footprint includes a triggering sound footprint based on one or more audio characteristics of the audio footprint. The system may detect synthetic sound based on the audio footprint and after determining that the audio footprint includes the triggering sound footprint, wherein the synthetic sound is indicative of a sound recording. The system may transmit a notification to one or more devices associated with the call based on detecting the synthetic sound.

Abstract: Systems, apparatuses, and methods are described for adjustment of card configurations for flight interruptions. Past flight interruptions that users experienced may be determined. Transactions conducted via the users' payment cards during the past flight interruptions may be determined. The data may be used for determining customized metrics for adjusting card configurations of the payment cards during current flight interruptions. Default metrics for adjusting card configurations of the payment cards during current flight interruptions may be determined. Current flight statuses may be monitored for the users. Current flight interruptions may be detected for the users. Card configurations of the payment cards may be adjusted during the detected current flight interruptions.

Abstract: A system and method for allowing a single live customer service agent to simultaneously serve multiple customers. According to various embodiments, a virtual agent operates at the front end to receive and attempt to handle customer issues. The virtual agent employs speech recognition and intent mapping in order to generate a proposed response that attempts to identify and resolve customer issues. In some scenarios, the proposed response includes both a response message and a response action to be taken. A chat history and the proposed response is then provided to the live agent. The live agent reviews the information provided, and determines whether the proposed response is appropriate. The live agent then approves the response or revises or replaces the response. The final response is sent back to the virtual agent for processing and providing to the customer.

Abstract: In some implementations, a system may capture audio from a call between a calling device and a called device. The system may filter the captured audio to generate a background audio layer. The system may generate an audio footprint that is a representation of sound in the background audio layer. The system may determine that the audio footprint includes a triggering sound footprint based on one or more audio characteristics of the audio footprint. The system may detect synthetic sound based on the audio footprint and after determining that the audio footprint includes the triggering sound footprint, wherein the synthetic sound is indicative of a sound recording. The system may transmit a notification to one or more devices associated with the call based on detecting the synthetic sound.

Abstract: Systems and methods of the present disclosure enable database search. The systems and/or methods may include receiving a search query that includes an input document having text. Word embeddings are generated within the input document, where the word embeddings include vector representations of words in the text of the input document. An average input document word embedding vector is determined for the word embeddings of the input document. A set of stored documents is accessed, where each stored document includes a stored text has a particular average stored document word embedding vector. A similarity model is used to determine a similarity metric measuring the similarity between the input document and each stored document based on the average input document word embedding vector and the particular average stored document word embedding vector of each stored document.

Abstract: A system may include processor(s), and memory in communication with the processor(s) and storing instructions configured to cause the system to correct ASR errors. The system may receive a transcription comprising transcribed word(s) and may determine whether the transcribed word(s) exceed associated predefined confidence level(s). Responsive to determining a transcribed word does not exceed a predefined confidence level, the system may generate a predicted word. The system may calculate a distance between numerical representations of the transcribed word and the predicted word and may determine whether the distance exceeds a predefined threshold. Responsive to determining the distance exceeds the predefined threshold, the system may determine whether at least one red flag word of a list of red flag words corresponds to a context of the transcription, and, responsive to making that determination, may classify the transcription as associated with a first category.

Abstract: In some implementations, a credit decision platform may receive a credit request from an applicant and obtain domestic historical data associated with the applicant from a credit bureau device. The credit decision platform may obtain access to an email account associated with the applicant based on determining that the domestic historical data associated with the applicant is insufficient to process the credit request. The credit decision platform may identify, using one or more machine learning models, a set of email messages included in the email account that are relevant to the credit request and may analyze content included in the set of email messages to generate non-domestic historical data associated with the applicant. The credit decision platform may generate a decision on the credit request based on an estimated creditworthiness of the applicant, which may be determined based on the non-domestic historical data.

Abstract: A system may include processor(s), and memory in communication with the processor(s) and storing instructions configured to cause the system to correct ASR errors. The system may receive a transcription comprising transcribed word(s) and may determine whether the transcribed word(s) exceed associated predefined confidence level(s). Responsive to determining a transcribed word does not exceed a predefined confidence level, the system may generate a predicted word. The system may calculate a distance between numerical representations of the transcribed word and the predicted word and may determine whether the distance exceeds a predefined threshold. Responsive to determining the distance exceeds the predefined threshold, the system may determine whether at least one red flag word of a list of red flag words corresponds to a context of the transcription, and, responsive to making that determination, may classify the transcription as associated with a first category.

Abstract: Systems and methods of the present disclosure enable variable and dynamic feedback to electronic activities by receiving event data and utilizing a feedback machine learning model to predict an average feedback attribute and an average feedback variability attribute based on the event data. A feedback data entry is added to a user profile to store the average feedback attribute and the average feedback variability attribute. A new event and a new event attribute are received for the user profile, and a feedback probability distribution for the new event is generated based on the new event attribute, the average feedback attribute, and the average feedback variability attribute. A new event feedback is generated for the new event based on a random selection from the feedback probability distribution and is output to a computing device associated with the user profile.

Abstract: In some implementations, a credit decision platform may receive a credit request from an applicant and obtain domestic historical data associated with the applicant from a credit bureau device. The credit decision platform may obtain access to an email account associated with the applicant based on determining that the domestic historical data associated with the applicant is insufficient to process the credit request. The credit decision platform may identify, using one or more machine learning models, a set of email messages included in the email account that are relevant to the credit request and may analyze content included in the set of email messages to generate non-domestic historical data associated with the applicant. The credit decision platform may generate a decision on the credit request based on an estimated creditworthiness of the applicant, which may be determined based on the non-domestic historical data.

Abstract: Methods and systems disclosed herein describe customizing searching. Search queries may be customized according to a user's preferences. A user may emphasize or indicate that additional weight should be given to one or more terms in a search query. Terms that are weighted higher may have a larger impact on the results that are returned in response to the search query. In addition to changing the terms in a search query, a user may provide a weight for each term. Each term in a search query may be weighted to varying degrees, giving a user more control over the results that are returned. The weights may be used with machine learning techniques to generate a vector representation of a search query. The vector representation of the search query may be compared with vector representations of search objects to determine results that match the search query.