Dynamic speech resource allocation
View Patent ↗A call is received at an interactive voice response (IVR) system. A voice communications session is established between the IVR system and the telephonic device. A request from the IVR system to allocate a speech resource for processing voice data of the voice communications session is received by a dynamic speech allocation (DSA) engine. Configuration data associated with a current state of the voice communications session is accessed by the DSA engine. Dynamic characteristics associated with the caller are accessed by the DSA engine. A speech resource from among multiple speech resources is selected by the DSA engine based on the current state and the dynamic characteristics. The selected speech resource is allocated to the voice communications session by enabling the IVR system to use the selected speech resource to process voice data received from the caller during the current state of the voice communications session.
1. A computer-implemented method comprising:
receiving, by a call handling system, a request to allocate a speech resource for processing voice data of a voice communications session between an interactive voice response (IVR) system and a telephonic device;
accessing, by the call handling system, configuration data associated with a current state of the voice communications session;
determining, by the call handling system, one or more data processing requirements of the current state of the voice communications session;
selecting, by the call handling system, a selected speech resource from among multiple speech resources, each of the speech resources having an associated cost, at least two of the associated costs being different, the selecting being based on the configuration data, the one or more data processing requirements of the current state of the voice communications session, and the associated costs of the speech resources, the multiple speech resources comprising at least one automatic speech recognition (ASR) engine; and
allocating the selected speech resource to the voice communications session.
2. The method of claim 1 , comprising:
accessing, by the call handling system, dynamic interaction data associated with a user of the telephonic device.
3. The method of claim 2 , wherein the dynamic interaction data includes data representing one or more voice characteristics associated with the user.
4. The method of claim 2 , wherein the dynamic interaction data includes data representing characteristics associated with the user's calling environment during the voice communications session.
5. The method of claim 2 , wherein the dynamic interaction data includes a location of the user during the current state of the voice communications session.
6. The method of claim 1 , wherein selecting a speech resource comprises selecting at least one automatic speech recognition (ASR) engine based on one or more ASR engine attributes, and wherein the one or more ASR engine attributes include a speech type, a supported language, a channel type, a cost per transaction, a recognition accuracy, a security feature, or an interaction type.
7. The method of claim 1 ,
comprising accessing, by the call handling system, interaction data associated with a previous voice communications session; and
wherein selecting, by the call handling system, a speech resource from among multiple speech resources further comprises selecting the speech resource based on the configuration data and the interaction data.
8. The method of claim 1 , comprising:
determining that the selected speech resource does not satisfy a demand of a user of the telephonic device; and
in response to determining that the selected speech resource does not satisfy the demand of the user, selecting, by the call handling system, a second, different, speech resource from among the multiple speech resources; and
allocating the second speech resource to the voice communications session.
9. The computer-implemented method of claim 1 , wherein the one or more data processing requirements of the current state of the voice communications session comprises an ambient noise level of the voice communications session.
10. A system comprising:
one or more computers and one or more storage devices storing instructions that when executed by the one or more computers cause the one or more computers to perform operations comprising:
receiving a request to allocate a speech resource for processing voice data of a voice communications session between an interactive voice response (IVR) system and a telephonic device;
accessing configuration data associated with a current state of the voice communications session;
determining one or more data processing requirements of the current state of the voice communications session;
selecting a speech resource from among multiple speech resources, each of the speech resources having an associated cost, at least two of the associated costs being different, the selecting being based on the configuration data, the one or more data processing requirements of the current state of the voice communications session, and the associated costs of the speech resources, the multiple speech resources comprising at least one automatic speech recognition (ASR) engine; and
allocating the selected speech resource to the voice communications session.
11. The system of claim 10 , comprising:
accessing dynamic interaction data associated with a user of the telephonic device.
12. The system of claim 11 , wherein the dynamic interaction data includes (i) data representing one or more voice characteristics associated with the user, (ii) data representing characteristics associated with the user's calling environment during the voice communications session, or (iii) data representing a location of the user during the current state of the voice communications session.
13. The system of claim 10 , wherein selecting a speech resource comprises selecting one of the at least one automatic speech recognition (ASR) engine based on one or more ASR engine attributes, and wherein the one or more ASR engine attributes include a speech type, a supported language, a channel type, a cost per transaction, a recognition accuracy, a security feature, or an interaction type.
14. The system of claim 10 ,
wherein the operations comprise accessing interaction data associated with a previous voice communications session; and
wherein selecting a speech resource from among multiple speech resources further comprises selecting the speech resource based on the configuration data and the interaction data.
15. The system of claim 10 , wherein the operations comprise:
determining that the selected speech resource does not satisfy a demand of a user of the telephonic device; and
in response to determining that the selected speech resource does not satisfy the demand of the user, selecting, a second, different, speech resource from among the multiple speech resources; and
allocating the second speech resource to the voice communications session.
16. A non-transitory computer-readable medium storing software having stored thereon instructions, which, when executed by one or more computers, cause the one or more computers to perform operations of:
receiving a request to allocate a speech resource for processing voice data of a voice communications session between an interactive voice response (IVR) system and a telephonic device;
accessing configuration data associated with a current state of the voice communications session;
determining one or more data processing requirements of the current state of the voice communications session;
selecting a speech resource from among multiple speech resources, each of the speech resources having an associated cost, at least two of the associated costs being different, the selecting being based on the configuration data, the one or more data processing requirements of the current state of the voice communications session, and the associated costs of the speech resources, the multiple speech resources comprising at least one automatic speech recognition (ASR) engine; and
allocating the selected speech resource to the voice communications session.
17. The non-transitory computer-readable medium of claim 16 , comprising:
accessing dynamic interaction data associated with a user of the telephonic device.
18. The non-transitory computer-readable medium of claim 17 , wherein the dynamic interaction data includes (i) data representing one or more voice characteristics associated with the user, (ii) data representing characteristics associated with the user's calling environment during the voice communications session, or (iii) data representing a location of the user during the current state of the voice communications session.
19. The non-transitory computer-readable medium of claim 16 , wherein selecting a speech resource comprises selecting one of the at least one automatic speech recognition (ASR) engine based on one or more ASR engine attributes, and wherein the one or more ASR engine attributes include a speech type, a supported language, a channel type, a cost per transaction, a recognition accuracy, a security feature, or an interaction type.
20. The non-transitory computer-readable medium of claim 16 ,
wherein the operations comprise accessing interaction data associated with a previous voice communications session; and
wherein selecting a speech resource from among multiple speech resources further comprises selecting the speech resource based on the configuration data and the interaction data.
21. The non-transitory computer-readable medium of claim 16 , wherein the operations comprise:
determining that the selected speech resource does not satisfy a demand of a user of the telephonic device; and
in response to determining that the selected speech resource does not satisfy the demand of the user, selecting, a second, different, speech resource from among the multiple speech resources; and
allocating the second speech resource to the voice communications session.