Cluster based CRM
In a crowd sourcing approach, responses to customer service inquiries are provided by routing a subset of the inquiries to an independent group of experts. The customer service inquiries are optionally routed to specific experts based on matches between identified subject matter of the inquiries and expertise of the experts. Embodiments include an expert system configured to identify clusters of inquiries that can be responded to using predetermined response content. The expert system is optionally trained based on scored responses to prior customer service inquiries.
1. An expert management system configured to manage customer service inquiries, the system comprising:
a request I/O configured to receive a stream of customer service inquiries;
status logic configured to monitor statuses of a plurality of human experts;
expert data storage configured to store data characterizing one or more specialty characteristics of each of the plurality of experts;
cluster detection logic configured to assign the customer service inquiries to clusters of the customer service requests, the clusters having common characteristics or being satisfied using similar responses;
matching logic to match individual human experts to particular members of the customer service inquiries, the matching being at least partially based on membership of the particular customer service request being in particular members of the identified clusters; and
routing logic configured to route the customer service inquiries to members of the human experts.
2. The system of claim 1 , further comprising automated response logic configured to provide at least partial responses to the customer service inquiries.
3. The system of claim 2 , wherein the matching logic includes a first machine learning system configured to predict an ability of the automated response logic to provide satisfactory responses to particular members of the customer service inquires and to predict an ability of the human experts to provide satisfactory responses to particular members of the customer service inquiries.
4. The system of claim 3 , wherein the first machine learning system is trained based on scores of responses to customer service inquiries.
5. The system of claim 4 , wherein the scores are generated using customer feedback.
6. The system of claim 4 , wherein the scores are generated using input from members of the human experts.
7. The system of claim 2 , wherein the automated response logic includes a second machine learning system configured to provide the at least partial response to members of the customer service inquiries.
8. The system of claim 1 , wherein the cluster detection logic is configured to identify new clusters of customer service inquiries, the new cluster having different characteristics relative to prior clusters of customer service inquiries.
9. The system of claim 1 , wherein the cluster detection logic includes a machine learning system trained based on customer service inquiries and scored responses to these customer service inquiries.
10. An expert management system configured to manage customer service inquiries, the system comprising:
a request I/O configured to receive a stream of customer service inquiries;
status logic configured to monitor statuses of a plurality of human experts;
expert data storage configured to store data characterizing one or more specialty characteristics of each of the plurality of experts;
automated response logic configured to provide at least partial responses to the customer service inquiries;
cluster matching logic configured to assign members of the customer service inquiries to clusters of customer services inquires, the clusters including common topics and requiring similar responses;
matching logic including a first machine learning system configured to predict an ability of the automated response logic to provide satisfactory responses to particular members of the customer service inquires and to predict an ability of human experts to provide satisfactory responses to particular members of the customer service inquiries, wherein the predicted ability is at least partially based on inclusion of particular members of the customer service inquires in an cluster; and
routing logic configured to route the customer service inquiries to members of the human experts or the automated response logic, in the alternative, based on the predicted ability.
11. The system of claim 10 , wherein the common topics include password help, balance inquiries, reservations and setup for a specific device.
12. The system of claim 10 , wherein the prediction that of the ability of the human experts to provide satisfactory responses are based on response scores of the human experts for a particular cluster of customer service inquiries.
13. The system of claim 10 , further comprising training logic configured to train the cluster identification logic.
14. The system of claim 13 , wherein the cluster identification logic is trained using response scores.
15. The system of claim 10 , wherein the cluster matching logic is configured to detect new clusters of customer services inquires, the new clusters including common topics and requiring similar responses.
16. The system of claim 10 , wherein the cluster matching logic is configured to assign the members of the customer service inquiries to the clusters of customer services inquires based on an analysis of text within each of the customer service inquiries.
17. The system of claim 10 , wherein the cluster matching logic is configured to assign the members of the customer service inquiries to the clusters of customer services inquires based on metadata associated with each of the customer service inquiries.