US 12,431,140 B1
Load testing platform for voice assistant-enabled technologies
Puneeth Simha Kadaba Sathya Kumar, Pleasanton, CA (US); Saso Crnugelj-Gale, Half Moon Bay, CA (US); Chenyuan Wang, Sunnyvale, CA (US); Shilpi Nair, San Ramon, CA (US); Diyu Zhu, Milpitas, CA (US); and Arwin Submaraniam, Sammamish, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Jun. 30, 2022, as Appl. No. 17/855,365.
Int. Cl. G10L 15/34 (2013.01); G10L 15/30 (2013.01)
CPC G10L 15/30 (2013.01) [G10L 15/34 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
receiving, via a graphical user interface (GUI) provided by a load testing service for voice assistant-enabled devices, input specifying configurations for a load testing plan, wherein the configurations include:
an identifier of an execution environment including one or more compute instances to be used to launch emulators for voice assistant-enabled devices,
values indicating a distribution of two or more types of voice assistant-enabled devices to be launched in the execution environment,
one or more text utterances to be used to invoke interactions with the emulators for voice assistant-enabled devices, and
a rate of interactions to be invoked against the emulators for voice assistant-enabled devices in the execution environment;
launching, in the execution environment, the emulators based on the load testing plan, wherein the emulators include at least a first set of emulators for a first type of voice assistant-enabled device and a second set of emulators for a second type of voice assistant-enabled device that is different from the first type of voice assistant-enabled device, and wherein a number of the first set of emulators and the second set of emulators is based on the values indicating the distribution of two or more types of voice assistant-enabled devices against which to invoke interactions;
invoking interactions with the emulators based on audio data corresponding to the one or more text utterances and the rate of interactions to be invoked, wherein invoking the interactions involves the emulators interfacing with at least one service;
obtaining, from a monitoring service, a performance metric associated with the emulators or the at least one service; and
causing display of the performance metric.