Abstract: A method includes providing audio signals of an interaction between a plurality of human speakers, the speakers speaking into electronic devices to record the audio signals. The audio signals, which are optionally combined, include agent audio and subject audio. The method further includes automatically processing the audio signals to generate a speaker separated natural language transcript of the interaction from the audio signals. For each identified question, a subject response is identified. From the agent text, it is determined whether the question asked by the at least one agent is an open question or a closed question. A decision engine is used to determine the veracity of the subject response and the subject response is flagged if the indicia of the likelihood of deception in the subject response exceeds a predetermined value.

Abstract: The present invention provides a system for self-learning end to end automatic speech recognition. The system comprises a memory. The memory stores a language model, a noise database and an automatic speech recognition, ASR, model. The system further comprises a processor. The processor may be hardware based or it may be cloud based. The processor is configured to run: an application programming interface (API); a text extractor; a language model builder; a pronunciation model; a text to speech engine; and a mixer. The text to speech engine may also be referred to as a text to speech converter. The system provides automatic updating of the ASR model to improve accuracy of the ASR functionality of the system without the need for complete retraining.

Abstract: A database is stored on a first cloud server to store encrypted files. A table stored at a second cloud server is a key value store, wherein the keys are unique secret values and the values are encrypted database values. The unique secret values relate to a specific instance of an update operation of a specific keyword. An update operation may be adding a new file to the database or deleting a file from the database. Each unique secret value is associated with an encrypted database value. A file identifier can be extracted from the encrypted database value. A file identifier identifies the location of a file in the database in the first cloud server. Each time a file containing a keyword is added or deleted from the database, a new unique secret value is generated and added to the table along with its associated encrypted database value.

Abstract: A database is stored on a first cloud server to store encrypted files. A table stored at a second cloud server is a key value store, wherein the keys are unique secret values and the values are encrypted database values. The unique secret values relate to a specific instance of an update operation of a specific keyword. An update operation may be adding a new file to the database or deleting a file from the database. Each unique secret value is associated with an encrypted database value. A file identifier can be extracted from the encrypted database value. A file identifier identifies the location of a file in the database in the first cloud server. Each time a file containing a keyword is added or deleted from the database, a new unique secret value is generated and added to the table along with its associated encrypted database value.

Abstract: A method includes providing audio signals of an interaction between a plurality of human speakers, the speakers speaking into electronic devices to record the audio signals. The audio signals, which are optionally combined, include agent audio and subject audio. The method further includes automatically processing the audio signals to generate a speaker separated natural language transcript of the interaction from the audio signals. For each identified question, a subject response is identified. From the agent text, it is determined whether the question asked by the at least one agent is an open question or a closed question. A decision engine is used to determine the veracity of the subject response and the subject response is flagged if the indicia of the likelihood of deception in the subject response exceeds a predetermined value.

Abstract: The present invention provides a secure method for outsourcing data analysis to a third party without the third party being able to access the data in clear. In the present invention, the client’s data is never written to disk, this feature provides the security. The data is decrypted and processed in the enclave which ensures that the data is secure and hidden at all times when it is visible “in clear”.

Abstract: A method includes providing audio signals of an interaction between a plurality of human speakers, the speakers speaking into electronic devices to record the audio signals. The audio signals, which are optionally combined, include agent audio and subject audio. The method further includes automatically processing the audio signals to generate a speaker separated natural language transcript of the interaction from the audio signals. For each identified question, a subject response is identified. From the agent text, it is determined whether the question asked by the at least one agent is an open question or a closed question. A decision engine is used to determine the veracity of the subject response and the subject response is flagged if the indicia of the likelihood of deception in the subject response exceeds a predetermined value.