Systems and methods for interactive voice response auto-scaling
Systems and methods for distributing discrete events into a stateless cloud provider architecture are disclosed. Certain embodiments involve receiving, by the load balancer, a scaling up indication for the IVR system, the scaling up indication comprising an identification of a new server to be added to the plurality of servers. The method involves responsive to receiving the scaling up indication, registering the new server for IVR traffic by the load balancer. The method includes after registering the new server, receiving an IVR event. The method includes distributing the IVR event to the new server.
1. A method for operating a scalable interactive voice response (IVR) system, the IVR system comprising a plurality of IVR servers, the method comprising:
receiving an IVR event;
retrieving a plurality of historical scaling events;
deriving at least one session parameter from the plurality of historical scaling events;
determining a predicted session requirement based on the received IVR event and on the derived at least one session parameter; and
assigning the IVR event to one of the plurality of IVR servers based on the predicted session requirement.
2. The method of claim 1 , wherein the IVR event represents an interaction between the IVR system and a user of a mobile device.
3. The method of claim 1 , wherein the at least one session parameter from comprises one or more of:
a timestamp of an historical scaling event of the plurality of historical scaling events;
a scaling factor of an historical scaling event of the plurality of historical scaling events; or
a direction of scaling for an historical scaling event of the plurality of historical scaling events.
4. The method of claim 1 , wherein the determining the predicted session requirement utilizes machine learning.
5. The method of claim 1 , wherein:
the IVR system further comprises a load balancer;
the predicted session requirement comprises determining that a capacity of the plurality of IVR servers is less than a predicted number of sessions, and
the method further comprises:
in response to the capacity of the plurality of IVR servers being less than the predicted number of sessions, generating a scaling up indication for the IVR system, the scaling up indication comprising an identification of a new server to be added to the plurality of servers; and
responsive to receiving the scaling up indication, registering the new server for IVR traffic by the load balancer.
6. The method of claim 5 , wherein the assigning the IVR event to one of the plurality of IVR servers comprises assigning the IVR event to the new server.
7. The method of claim 5 , wherein the load balancer communicates the scaling up indication to a stateless cloud platform and the IVR system is a stateful system.
8. The method of claim 1 , wherein:
determining the predicted session requirement comprises determining that a capacity of the plurality of IVR servers exceeds a predicted number of sessions, and
the method further comprises:
in response to determining that the capacity of the plurality of IVR servers exceeds the predicted number of sessions, generating a scaling down indication for the IVR system, the scaling down indication comprising an identification of a selected server to be removed from the plurality of servers;
responsive to generating the scaling down indication, de-registering the selected server from IVR traffic by the load balancer;
after de-registering the selected server, preventing assignment of the IVR event to the selected server; and
communicating, by the IVR system to a cloud provider, a request for termination of the selected server.
9. The method of claim 8 , wherein the IVR system further comprises a load balancer, wherein de-registering the selected server comprises:
removing from a list of available servers, by the load balancer, the selected server with one or more of a virtual machine ID, a cloud service provider instance identifier, or other unique identifier.
10. A method for operating a scalable interactive voice response (IVR) system, the IVR system comprising a plurality of IVR servers and a load balancer, the method comprising:
retrieving a plurality of historical scaling events;
deriving at least one session parameter from the plurality of historical scaling events;
determining that a capacity of the plurality of IVR servers is less than a predicted number of sessions on the derived at least one session parameter;
in response to the capacity of the plurality of IVR servers being less than the predicted number of sessions, generating a scaling up indication for the IVR system, the scaling up indication comprising an identification of a new server to be added to the plurality of servers;
responsive to receiving the scaling up indication, registering the new server for IVR traffic by the load balancer;
receiving an IVR event; and
assigning the IVR event to the new server.
11. The method of claim 10 , wherein the IVR event represents an interaction between the IVR system and a user of a mobile device.
12. The method of claim 10 , wherein the at least one session parameter comprises one of more of:
a timestamp of an historical scaling event of the plurality of historical scaling events;
a scaling factor of an historical scaling event of the plurality of historical scaling events; or
a direction of scaling for an historical scaling event of the plurality of historical scaling events.
13. The method of claim 10 , wherein the determining the predicted number of sessions utilizes machine learning.
14. The method of claim 10 , wherein the load balancer communicates the scaling up indication to a stateless cloud platform and the IVR system is a stateful system.
15. The method of claim 14 , wherein the cloud platform initiates a stateless scaling event and wherein the interactive voice response system adjusts a number of discrete available sessions from a cloud platform capacity.
16. A method for operating a scalable interactive voice response (IVR) system, the IVR system comprising a plurality of IVR servers and a load balancer, the method comprising:
retrieving a plurality of historical scaling events;
deriving at least one session parameter from the plurality of historical scaling events;
determining that a capacity of the plurality of IVR servers exceeds a predicted number of sessions on the derived at least one session parameter;
in response to the capacity of the plurality of IVR servers being less than the predicted number of sessions, generating a scaling down indication for the IVR system, the scaling down indication comprising an identification of a selected server to be removed from the plurality of servers;
responsive to generating the scaling down indication, de-registering the selected server from IVR traffic by the load balancer;
after de-registering the selected server, preventing assignment of an IVR event to the selected server; and
communicating, by the IVR system to a cloud provider, a request for termination of the selected server.
17. The method of claim 16 , wherein de-registering the selected server comprises:
removing from a list of available servers, by the load balancer, the selected server with one or more of a virtual machine ID, a cloud service provider instance identifier, or other unique identifier.
18. The method of claim 16 , wherein the determining the predicted number of sessions utilizes machine learning.
19. The method of claim 16 , wherein the load balancer communicates the scaling down indication to a stateless cloud platform and the IVR system is a stateful system.
20. The method of claim 19 , wherein the cloud platform initiates a stateless scaling event and wherein the interactive voice response system adjusts a number of discrete available sessions from a cloud platform capacity.