Selective aggregation of enqueued nodes into a subnetwork for simultaneous communication
Contact centers often encounter situations where a group of customers have a common issue to be addressed. Connecting each customer to an agent for one-on-one communication may be effective, but it is resource intensive both in terms of networking and computer requirements and the time an agent spends repeating the same information. By performing an initial assessment of customers' issues, a common issue may be discovered that a common issue exists that can be handled in a group lecture pod, wherein a number of customer connect to interact with an agent and optionally to each other. Provided confidential information is not required, and any other required consent given, customers may be connected to the group lecture pod for more immediate resolution of a common issue and not have to stay in queue to wait for an agent to become available for a one-on-one interaction.
1. A system, comprising:
a processor configured with machine-readable instructions maintained in a non-transitory memory; and
an interface to a network for simultaneous communication with nodes via the network;
wherein the processor performs:
obtaining issues to be addressed from each of the nodes;
determining, from the issues to be addressed from each of the nodes, whether a common issue exists that is common to a subset of the nodes;
evaluating whether the common issue is resolvable by a single agent simultaneously engaged with each of a plurality of the subset of the nodes;
upon determining the common issue is resolvable by the single agent, establishing a subnetwork comprising each of the plurality of the subset of nodes and a first agent node of the single agent for simultaneous communication thereon;
determining an excluded node of the nodes exists, wherein the excluded node was not added to the subnetwork; and
enqueuing the excluded node for connection with a second agent node of a second agent for one-on-one communication therebetween.
2. The system of claim 1 , wherein the processor performs determining whether the common issue exists that is common to the subset of the nodes, further comprising the processor determining:
a first estimated handling time to address each issue for the plurality of the subset of nodes utilizing one-on-one communication with one of a pool of agents, each utilizing a corresponding one of a pool of agent nodes; and
a second estimated handling time to address each issue for the plurality of the subset of nodes utilizing the subnetwork, and
wherein the common issue exists upon determining the second estimated handling time is less than the first estimated handling time.
3. The system of claim 2 , wherein the first estimated handling time further comprises a time enqueued by each of the nodes before the one-on-one communication is established.
4. The system of claim 1 , wherein the processor performs the determining of whether common issues exist comprising:
prompting a user associated with each of the nodes to describe an issue;
receiving from each of the nodes a description of the issue; and
providing each description of the issue as a first input to a neural network trained to determine common issues and receiving a decision therefrom.
5. The system of claim 4 , wherein the neural network is trained to determine common issues, comprising:
accessing a set of past issues;
applying one or more transformations to each past issue of the set of past issues including one or more of substituting a word with a synonymous word, substituting a word with a synonymous phrase, substituting a problem for an object associated with the problem, substituting the object for the problem associated with the object, inserting at least one redundant word, or removing at least one redundant word to create a modified set of past issues;
creating a first training set comprising the set of past issues, the modified set of past issues, and a set of unique issues;
training the neural network in a first state using the first training set;
creating a second training set for a second state of training comprising the first training set and the set of unique issues that are incorrectly identified as being common issues; and
training the neural network in the second state of training using the second training set.
6. The system of claim 1 , wherein the processor determines, from the issues to be addressed from each of the nodes, whether the common issue exists that is common to the subset of the nodes, further comprising determining whether the common issue exists that does not require knowledge of information previously determined to be private in order to select a resolution to address the common issue.
7. The system of claim 1 , wherein the processor determines, from the issues to be addressed from each of the nodes, whether the common issue exists that is common to the subset of the nodes, further comprising determining whether the common issue exists that does not require providing information previously determined to be private in order to carry out a resolution to address the common issue.
8. The system of claim 1 , further comprising the processor:
accessing a queue comprising an ordered number of entries, each entry comprising an indicia of a corresponding node of the nodes and wherein a qualified agent becomes available accessing the entry in a first position within the ordered number of entries is accessed and the corresponding node evaluated for connection to an agent node of the qualified agent;
wherein the evaluation comprises determining whether the corresponding node is a member of the subnetwork;
upon determining the corresponding node is not a member of the subnetwork, the evaluation is determined in the affirmative and the corresponding node connected to the agent node, the entry in the first position is removed from the queue, and each remaining entry is moved up in the queue;
upon determining the corresponding node is a member of the subnetwork, the evaluation is determined as a user response to a prompt to the corresponding node offering to disconnect from the subnetwork and connect to the qualified agent for a one-on-one interaction; and
upon the user response being in the affirmative, disconnecting the corresponding node from the subnetwork and connecting to the qualified agent.
9. The system of claim 8 , wherein the processor, upon the user response being in the negative, maintains the corresponding node in the subnetwork and moves the entry in the first position within the ordered number of entries to a second position within the ordered number of entries.
10. A method, comprising:
accessing nodes;
obtaining issues to be addressed from each of the nodes;
determining, from the issues to be addressed from each of the nodes, whether a common issue exists that is common to a subset of the nodes;
evaluating whether the common issue is resolvable by a single agent simultaneously engaged with each of a plurality of the subset of the nodes;
upon determining the common issue is resolvable by the single agent, establishing a subnetwork comprising each of the plurality of the subset of nodes and a first agent node of the single agent for simultaneous communication thereon;
determining an excluded node of the nodes exists, wherein the excluded node was not added to the subnetwork; and
enqueuing the excluded node for connection with a second agent node of a second agent for one-on-one communication therebetween.
11. The method of claim 10 , further comprising:
obtaining an issue to be addressed from a new node;
determining whether the issue to be addressed is sufficiently similar to the common issue; and
upon determining the issue to be addressed is sufficiently similar to the common issue, adding the new node to the subnetwork for simultaneous communication thereon.
12. The method of claim 10 , wherein determining whether the common issue exists that is common to the subset of the nodes, further comprises a processor determining:
a first estimated handling time to address each issue for the plurality of the subset of nodes utilizing one-on-one communication with one of a pool of agents, each utilizing a corresponding one of a pool of agent nodes; and
a second estimated handling time to address each issue for the plurality of the subset of nodes utilizing the subnetwork, and
wherein the common issue exists upon determining the second estimated handling time is less than the first estimated handling time.
13. The method of claim 10 , wherein determining, from the issues to be addressed from each of the nodes, whether the common issue exists that is common to the subset of the nodes, further comprising determining whether the common issue exists that does not require knowledge of information previously determined to be private in order to perform at least one of selecting of a resolution to address the common issue or to carry out the resolution to address the common issue.
14. A method, comprising:
accessing nodes;
obtaining issues to be addressed from each of the nodes;
determining, from the issues to be addressed from each of the nodes, whether a common issue exists that is common to a subset of the nodes;
evaluating whether the common issue is resolvable by a single agent simultaneously engaged with each of a plurality of the subset of the nodes; and
upon determining the common issue is resolvable by the single agent, establishing a subnetwork comprising each of the plurality of the subset of nodes and a first agent node of the single agent for simultaneous communication thereon; and
wherein determining whether the common issue exists that is common to the subset of the nodes, further comprises a processor determining:
a first estimated handling time to address each issue for the plurality of the subset of nodes utilizing a one-on-one communication with one of a pool of agents, each utilizing a corresponding one of a pool of agent nodes; and
a second estimated handling time to address each issue for the plurality of the subset of nodes utilizing the subnetwork; and
wherein the common issue exists upon determining the second estimated handling time is less than the first estimated handling time.
15. The method of claim 14 , wherein determining, from the issues to be addressed from each of the nodes, whether the common issue exists that is common to the subset of the nodes, further comprising determining whether the common issue exists that does not require knowledge of information previously determined to be private in order to perform at least one of selecting of a resolution to address the common issue or to carry out the resolution to address the common issue.
16. The method of claim 14 , further comprising:
obtaining an issue to be addressed from a new node;
determining whether the issue to be addressed is sufficiently similar to the common issue; and
upon determining the issue to be addressed is sufficiently similar to the common issue, adding the new node to the subnetwork for simultaneous communication thereon.
17. The method of claim 14 , wherein the first estimated handling time further comprises a time enqueued by each of the nodes before the one-on-one communication is established.
18. The method of claim 14 , wherein determining whether the common issue exist comprises:
prompting a user associated with each of the nodes to describe the common issue;
receiving from each of the nodes a description of the common issue; and
providing each description of the common issue as a first input to a neural network trained to determine common issues and receiving a decision therefrom.
19. The method of claim 18 , wherein the neural network is trained to determine common issues, comprising:
accessing a set of past issues;
applying one or more transformations to each past issue of the set of past issues including one or more of substituting a word with a synonymous word, substituting a word with a synonymous phrase, substituting a problem for an object associated with the problem, substituting the object for the problem associated with the object, inserting at least one redundant word, or removing at least one redundant word to create a modified set of past issues;
creating a first training set comprising the set of past issues, the modified set of past issues, and a set of unique issues;
training the neural network in a first state using the first training set;
creating a second training set for a second state of training comprising the first training set and the set of unique issues that are incorrectly identified as being common issues; and
training the neural network in the second state of training using the second training set.
20. The method of claim 14 , wherein the processor determines, from the issues to be addressed from each of the nodes, whether the common issue exists that is common to the subset of the nodes, further comprising determining whether the common issue exists that does not require knowledge of information previously determined to be private in order to select a resolution to address the common issue.