Agent instance live-monitoring by a management network for burnout and attrition prediction and response
A computer-implemented method for managing a contact center having a plurality of agents includes (a) receiving real-time data associated with each of a plurality of agents servicing incoming communications for the contact center, (b) categorizing each of the plurality of agents into a respective one of a plurality of burnout risk categories by processing the real-time data associated with each of the plurality of agents using a trained supervised machine learning model having a plurality of input features, (c) determining, based on a specification of a logical directive, an operation to be performed in relation to at least one agent in the plurality of agents, wherein the specification has at least one condition relating to the categorizing of each of the plurality of agents into the respective one of the plurality of burnout risk categories, and wherein the specification defines the operation to be performed in relation to the at least one agent in the plurality of agents upon the condition being satisfied, and (d) causing the operation to be performed in relation to the at least one agent instance in the plurality of agents. A computing system and article of manufacture are also provided.
1. A computing system for managing a contact center, the computing system comprising:
one or more hardware processors configured to execute instructions stored on at least one non-transitory computer readable medium to perform tasks including:
receiving real-time data associated with each of a plurality of agents servicing incoming communications for the contact center;
categorizing each of the plurality of agents into a respective one of a plurality of burnout risk categories by processing the real-time data associated with each of the plurality of agents using a trained supervised machine learning model having a plurality of input features;
determining, based on a specification of a logical directive, an operation to be performed in relation to at least one agent in the plurality of agents, wherein the specification has at least one condition relating to the categorizing of each of the plurality of agents into the respective one of the plurality of burnout risk categories, and wherein the specification defines the operation to be performed in relation to the at least one agent in the plurality of agents upon the condition being satisfied; and
causing the operation to be performed in relation to the at least one agent instance in the plurality of agents.
2. The computing system of claim 1 , wherein the trained supervised machine learning model is trained using historical data associated with the contact center.
3. The computing system of claim 2 , wherein the tasks further comprise, in order to train the trained supervised machine learning model:
calculating a correlation score for each of the plurality of input features used by the supervised machine learning model in the classifying of each of the plurality of agents, in order to determine how relatively informative each of the plurality of input features is for the classifying burnout risk; and
identifying at least one burnout risk predictor based on the calculated correlation score for each of the plurality of input features.
4. The computing system of claim 1 , wherein the trained supervised machine learning model is periodically retrained using historical data associated with the contact center.
5. The computing system of claim 1 , wherein the operation includes transmitting a message to the at least one agent and/or transmitting a message to at least one supervisor.
6. The computing system of claim 1 , wherein causing the operation to be performed includes sending an instruction to at least one of (a) a communication distributor server, (b) a workforce management server, or (c) a back-office case system server.
7. The computing system of claim 1 , wherein the operation is selected from the group consisting of (a) changing a state of the at least agent, (b) modifying an assigned schedule for the at least one agent, and (c) modifying an assigned queue of the at least one agent.
8. The computing system of claim 1 , wherein the operation includes displaying, on a display device, a graphical indication relating to burnout risk category for at least one of the plurality of agents.
9. The computing system of claim 8 , wherein the graphical indication is part of an interactive dashboard displayed on the display device.
10. The computing system of claim 1 , wherein the plurality of input features includes at least one input feature selected from the group consisting of (a) average handle time, (b) average time in after-call work, (c) average hold time, (d) average on-call time, and (e) average occupancy.
11. The computing system of claim 1 , wherein the received data associated with each of the plurality of agents includes respective agent instance data for each of the plurality of agents.
12. The computing system of claim 1 , wherein the supervised machine learning model is selected from the group consisting of a linear discriminant analysis (LDA) model, a quadratic discriminant analysis (QDA) model, a logistic regression model, and a survival analysis model.
13. A computer-implemented method for managing a contact center having a plurality of agents, the method comprising:
receiving, by at least one computing device, real-time data associated with each of a plurality of agents servicing incoming communications for the contact center;
categorizing, by the at least one computing device, each of the plurality of agents into a respective one of a plurality of burnout risk categories by processing the real-time data associated with each of the plurality of agents using a trained supervised machine learning model having a plurality of input features;
determining, by the at least one computing device, based on a specification of a logical directive, an operation to be performed in relation to at least one agent in the plurality of agents, wherein the specification has at least one condition relating to the categorizing of each of the plurality of agents into the respective one of the plurality of burnout risk categories, and wherein the specification defines the operation to be performed in relation to the at least one agent in the plurality of agents upon the condition being satisfied; and
causing, by the at least one computing device, the operation to be performed in relation to the at least one agent instance in the plurality of agents.
14. The computer-implemented method of claim 13 , wherein the trained supervised machine learning model is trained using historical data associated with the contact center.
15. The computer-implemented method of claim 14 , further comprising, in order to train the trained supervised machine learning model:
calculating, by the at least one computing device, a correlation score for each of the plurality of input features used by the supervised machine learning model in the classifying of each of the plurality of agents, in order to determine how relatively informative each of the plurality of input features is for the classifying burnout risk; and
identifying, by the at least one computing device, at least one burnout risk predictor based on the calculated correlation score for each of the plurality of input features.
16. The computer-implemented method of claim 13 , wherein the trained supervised machine learning model is periodically retrained using historical data associated with the contact center.
17. The computer-implemented method of claim 13 , wherein the operation includes transmitting a message to the at least one agent and/or transmitting a message to at least one supervisor.
18. The computer-implemented method of claim 13 , wherein causing the operation to be performed includes the at least one computing device sending an instruction to at least one of (a) a communication distributor server, (b) a workforce management server, or (c) a back-office case system server.
19. The computer-implemented method of claim 13 , wherein the operation is selected from the group consisting of (a) changing a state of the at least agent, (b) modifying an assigned schedule for the at least one agent, and (c) modifying an assigned queue of the at least one agent.
20. The computer-implemented method of claim 13 , wherein the operation includes displaying, on a display device, a graphical indication relating to burnout risk category for at least one of the plurality of agents.
21. The computer-implemented method of claim 20 , wherein the graphical indication is part of an interactive dashboard displayed on the display device.
22. The computer-implemented method of claim 13 , wherein the plurality of input features includes at least one input feature selected from the group consisting of (a) average handle time, (b) average time in after-call work, (c) average hold time, (d) average on-call time, and (e) average occupancy.
23. The computer-implemented method of claim 13 , wherein the received data associated with each of the plurality of agents includes respective agent instance data for each of the plurality of agents.
24. The computer-implemented method of claim 13 , wherein the supervised machine learning model is selected from the group consisting of a linear discriminant analysis (LDA) model, a quadratic discriminant analysis (QDA) model, a logistic regression model, and a survival analysis model.
25. An article of manufacture including a non-transitory computer-readable medium, having stored thereon program instructions that, upon execution by one or more processors in at least one computing device, cause the one or more processors to perform tasks comprising:
receiving real-time data associated with each of a plurality of agents servicing incoming communications for the contact center;
categorizing each of the plurality of agents into a respective one of a plurality of burnout risk categories by processing the real-time data associated with each of the plurality of agents using a trained supervised machine learning model having a plurality of input features;
determining, based on a specification of a logical directive, an operation to be performed in relation to at least one agent in the plurality of agents, wherein the specification has at least one condition relating to the categorizing of each of the plurality of agents into the respective one of the plurality of burnout risk categories, and wherein the specification defines the operation to be performed in relation to the at least one agent in the plurality of agents upon the condition being satisfied; and
causing the operation to be performed in relation to the at least one agent instance in the plurality of agents.
26. The article of manufacture of claim 25 wherein the operation is selected from the group consisting of (a) transmitting a message to the at least one agent, (b) transmitting a message to at least one supervisor, (c) sending an instruction to a communication distributor server, (d) sending an instruction to a workforce management server, or (e) sending an instruction to a back-office case system server.
27. The article of manufacture of claim 25 , wherein the operation is selected from the group consisting of (a) changing a state of the at least agent, (b) modifying an assigned schedule for the at least one agent, and (c) modifying an assigned queue of the at least one agent.
28. The article of manufacture of claim 25 , wherein the operation includes displaying, on a display device, a graphical indication relating to burnout risk category for at least one of the plurality of agents.
29. The article of manufacture of claim 25 , wherein the plurality of input features includes at least one input feature selected from the group consisting of (a) average handle time, (b) average time in after-call work, (c) average hold time, (d) average on-call time, and (e) average occupancy.
30. The article of manufacture of claim 25 , wherein the supervised machine learning model is selected from the group consisting of a linear discriminant analysis (LDA) model, a quadratic discriminant analysis (QDA) model, a logistic regression model, and a survival analysis model.