Methods and systems provided herein ensure that communication sessions such as phone calls over PSTNs can be reconstructed in the event that the session ends unintentionally. Methods and systems include detecting a call ended, analyzing data associated with the call, determining, based on the data, the call ended unintentionally, and reconstructing the call based on the determination the call ended unintentionally.

A method for testing the functional capability of an emergency call device of a transportation vehicle. In an emergency mode, at least one specific audio signal is generated for a vehicle passenger by the emergency call device and emitted to the interior of the transportation vehicle. For testing, the emergency call device is put into a testing mode in which the at least one audio signal intended for the emergency is generated and wherein, in the testing mode, the generated audio signal passes, within the vehicle-internal emergency call device, through a signal path which is modified compared with the emergency mode.

A method for testing the functional capability of an emergency call device of a transportation vehicle. In an emergency mode, at least one specific audio signal is generated for a vehicle passenger by the emergency call device and emitted to the interior of the transportation vehicle. For testing, the emergency call device is put into a testing mode in which the at least one audio signal intended for the emergency is generated and wherein, in the testing mode, the generated audio signal passes, within the vehicle-internal emergency call device, through a signal path which is modified compared with the emergency mode.

A system for centralized testing of web-based agent desktops has been devised. The invention uses a test control portal. The test control portal acts as the interface between the client interaction software systems testing system and analyst controlled test device, executes an extensive set of robust test directive commands with underlying routines to be used to specify test conditions without the use of programming ability on the part of the analyst, uses a robust set of report item and format choice designators to allow easy selection of a range of report content and styles.

An artificial intelligence (“AI”) system for tuning a machine learning interactive voice response system is provided. The AI system may perform analysis of outputs generated by the machine learning models. The AI system may determine an expected model output for a given test input. The AI system may determine accuracy, precision and recall scores for an actual output garneted in response to the test input. The system may determine performance metrics for interim outputs generated by individual machine learning models within the interactive voice response system. The AI system may replace malfunctioning models with replacement models.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for evaluating the quality of a communication session or of a communication path used for the communication session. One of the methods includes initiating a communication session between a first communications device and a second communications device, wherein initiating the communication session comprises routing session data for the communication session along a first communication path between the first communications device and the second communications device; generating, at the first communications device, a plurality of reference content samples; generating a recording of the communication session as received at a first destination along the first communication path; and evaluating a quality of the communication session or of the first communication path by comparing the plurality of reference content samples with the recorded communication session.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for evaluating the quality of a communication session or of a communication path used for the communication session. One of the methods includes initiating a communication session between a first communications device and a second communications device, wherein initiating the communication session comprises routing session data for the communication session along a first communication path between the first communications device and the second communications device; generating, at the first communications device, a plurality of reference content samples; generating a recording of the communication session as received at a first destination along the first communication path; and evaluating a quality of the communication session or of the first communication path by comparing the plurality of reference content samples with the recorded communication session.

A system for automated contact center agent thick client workstation testing. The invention uses a workstation test execution manager. The workstation test execution manager acts as the interface between the client interaction software systems testing system and analyst controlled test device, executes an extensive set of robust test directive commands with underlying routines to be used to specify test conditions without the use of programming ability on the part of the analyst, uses a robust set of report item and format choice designators to allow easy selection of a range of report content and styles.

A system and methods for reliable call recording testing and proprietary customer information retrieval, wherein recording may be both active and passive in nature, recording call audio from the moment a call begins such as to include dial tones, IVR interactions, and other non-conversation call contents that may generally be omitted in traditional recordings. These recordings may then be compared to similar recordings collected by traditional systems to “align” the recordings, identifying portions of the call that may have been omitted from one recording by using audio fingerprinting to match recordings up for improved analysis.

A system and methods for reliable call recording testing and proprietary customer information retrieval, wherein recording may be both active and passive in nature, recording call audio from the moment a call begins such as to include dial tones, IVR interactions, and other non-conversation call contents that may generally be omitted in traditional recordings. These recordings may then be compared to similar recordings collected by traditional systems to “align” the recordings, identifying portions of the call that may have been omitted from one recording by using audio fingerprinting to match recordings up for improved analysis.