System for closed loop decisionmaking in an automated care system
View Patent ↗There is disclosed a system functional for operating and updating an interactive voice response self care system which utilizes various types of models. The system is capable of improving the performance of the self care system through updating the models used in that system in a closed loop manner. The system is also potentially configured to utilize input from a human agent, such as a customer service representative, to which a customer call is transferred by a self care system.
1. A system comprising:
a) one or more processors; and
b) a computer memory containing one or more models utilized by a processor from the one or more processors configured to process customer interactions comprising:
i) one or more statements made by a customer; and
ii) one or more prompts played for said customer;
c) a computer-readable storage medium storing a set of computer executable instructions and coupled to at least one of the one or more processors, wherein the at least one processor coupled to the computer-readable storage medium is configured to execute instructions to:
i) create a data record comprising a set of context information related to one or more of said processed customer interactions;
ii) store the set of context information from said data record in said computer memory;
iii) utilize the set of context information stored in said computer memory to automatically create one or more model updates without human intervention; and
iv) automatically update one or more of said models using one or more of said model updates;
(v) make at least one recommendation regarding processing said customer interactions, wherein the recommendation is one of replay the prompt or transfer the customer to a live agent; and
wherein:
a) said one or more models comprises a statistical prediction model:
b) said processor is further configured to execute instructions to make said at least one recommendation to replay the prompt or transfer the customer based on said statistical prediction model.
2. The system of claim 1 further comprising an agent terminal configured to allow an agent to process said customer interactions;
wherein said customer interaction further comprises one or more statements made by the agent to the customer; and
wherein said at least one processor is further configured to execute instructions to:
a) create a second data record comprising information related to the processing of said one or more processed customer interactions by the agent;
b) determine if said data record and said second data record contain information relating to a single customer from a single processed customer interaction, wherein said single processed customer interaction relates to said data record and said second data record;
c) if said data record and said second data record contain information relating to said single customer from said single processed customer interaction, correlate said data record and said second data record;
d) store information from said second data record in said computer memory.
3. The system of claim 1 wherein said processor is further configured to execute instructions to make said at least one recommendation based on a means for evaluating options.
4. The system of claim 1 wherein said one or more model updates created by said processor comprises an update for said statistical prediction model, which model provides a prediction value for each of at least two recommendations.
5. The system of claim 1 wherein said processor is further configured to execute instructions to:
a) identify a set of actions;
b) perform an evaluation of desirability of an action from said set of actions;
wherein said processor is further configured to execute instructions to make at least one recommendation to execute the action from the set of actions based at least in part on said evaluation of desirability of the action from said set of actions, wherein the action is one of replay the prompt or transfer the customer to a live agent.
6. The system of claim 1 wherein said one or more models in said computer memory comprises a statistical language model, and wherein said processor is further configured to execute instructions to determine at least one word included in a statement from said one or more customer statements based on said statistical language model.
7. The system of claim 6 wherein said one or more models in said computer memory comprises a statistical language model; wherein said processor is further configured to determine that as least one word included in a statement from said on or more customer statement is based on said statistical language model; and wherein said one or more model updates comprises an update for said statistical language model.
8. The system of claim 1 wherein said processor is further configured to execute instructions to determine a meaning of a statement from said one or more customer statements.
9. The system of claim 8 wherein:
a) said one or more models in said computer memory comprises a semantic classification model;
b) said processor is further configured to execute instructions to determine said meaning of a statement from said one or more customer statements based on said semantic classification model.
10. The system of claim 9 wherein said one or more model updates comprises an update for said semantic classification model.
11. The system of claim 1 wherein said one or more model updates comprise one or more updated models, said updated models comprising statistical prediction models that provide a prediction value for each of at least two recommendations.
12. The system of claim 1 wherein:
a) said computer memory further contains:
i) a set of potential call dispositions, wherein each disposition correlates to a potential outcome of said customer interaction;
ii) for each potential call disposition from the set of potential call dispositions, an indication of the potential call disposition's desirability;
iii) a record of a disposition of said customer interaction; and
b) said processor is further configured to execute instructions to make at least one recommendation regarding processing said customer interaction based at least in part on the indication of the potential call disposition's desirability.
13. The system of claim 1 wherein said processor is further configured to execute instructions to automatically update one or more of said models using one or more of said model updates in response to a trigger, wherein the trigger is a usage milestone.
14. The system of claim 1 wherein said processor is further configured to execute instructions to coordinate processing of said customer interaction through the implementation of one or more actions.