Heuristic sales agent training assistant
A heuristic engine includes capabilities to collect an unstructured data set, such as a agent sales transaction record, and generate an agent training suggestion for subsequent action. Providing the heuristic engine with relevant data sets and a current context may allow determination of predicted future customer contexts and subsequent actions. The heuristic engine may learn from past data transactions and appropriate correlations with events and available data.
1. A computer-implemented method, executed with a processor, comprising:
identifying, with the processor and from a first memory,
an un-structured website transaction data record from a plurality of stored un-structured website transaction data records, the identified un-structured website transaction data record indicating questions presented by a customer, and
an un-structured agent transaction record from a plurality of stored un-structured agent transaction records, the identified un-structured agent transaction record indicating interactions of an agent with a training interface or a customer service interface;
receiving, with the processor, a natural language input from the agent using a human machine interface;
accessing, with the processor, a heuristic algorithm stored in a second memory;
executing the heuristic algorithm, with the processor, to determine:
a current context classifying the customer into a particular category based upon a product type, the heuristic algorithm determining the current context based at least in part on the identified unstructured website transaction data record, and
a training resource for providing additional training to the agent, the heuristic algorithm determining the training resource based at least in part on the identified un-structured agent transaction record, the natural language input, and the current context;
determining, with the processor and based on the agent being trained with the training resource, an indication of modified agent behavior; and
training, with the processor, the heuristic algorithm in the second memory using a determined correlation between the training resource and the indication of modified agent behavior.
2. The computer-implemented method of claim 1 , wherein the identified un-structured agent transaction record comprises a natural language data set.
3. The computer-implemented method of claim 1 , further comprising calculating the determined correlation as a correlation score based at least in part on the training resource and the indication of modified agent behavior.
4. The computer-implemented method of claim 1 , wherein the identified un-structured agent transaction record comprises past transactions related to at least one agent.
5. The computer-implemented method of claim 1 , wherein the first memory comprises an external transaction server, and the second memory comprises an external heuristic server.
6. The computer-implemented method of claim 1 , wherein the questions presented by the customer comprise questions regarding a product characterized by the product type.
7. The computer-implemented method of claim 1 , wherein the identified un-structured agent transaction record comprises a plurality of sales records and the natural language input from the agent comprises a question regarding a pending sales transaction.
8. A computer system comprising one or more processors and/or transceivers and configured to:
identify, from a first memory,
an un-structured website transaction data record from a plurality of stored un-structured website transaction data records, the identified un-structured website transaction data record indicating questions presented by a customer, and
an un-structured agent transaction record from a plurality of stored un-structured agent transaction records, the identified un-structured agent transaction record indicating interactions of an agent with a training or customer service interface;
receive a natural language input from the agent using a human machine interface;
access a heuristic algorithm stored in a second memory;
execute the heuristic algorithm, to determine:
a current context classifying the customer into a particular category based upon a product type, the heuristic algorithm determining the current context based at least in part on the identified website transaction data record, and
a training resource for providing additional training to the agent, the heuristic algorithm determining the training resource based at least in part, on the identified un-structured agent transaction record, the natural language input, and the current context;
determine, based on the agent being trained with the training resource, an indication of modified agent behavior; and
train the heuristic algorithm in the second memory using a determined correlation between the training resource and the indication of modified agent behavior.
9. The computer system of claim 8 , wherein the identified un-structured agent transaction record comprises a natural language data set.
10. The computer system of claim 8 , wherein the computer system is further configured to:
calculate the determined correlation as a correlation score based at least in part on the training resource and the indication of modified agent behavior.
11. The computer system of claim 8 , wherein the identified un-structured agent transaction record comprises past transactions related to at least one agent, and the questions presented by the customer comprise questions associated with a sale of a product characterized by the product type.
12. The computer system of claim 8 , wherein the first memory comprises an external transaction server and the second memory comprises an external heuristic server.
13. The computer system of claim 8 , wherein the plurality of un-structured transaction records are associated with a plurality of agents, and wherein the identified un-structured transaction record comprises real-time interaction data, past interaction data, and customer service data associated with the agent.
14. A non-transitory computer readable medium, comprising computer readable instructions that when executed by a processor cause the processor to perform acts comprising:
identifying, with the processor,
an un-structured website transaction data record from a plurality of stored un-structured website transaction data records, the identified un-structured website transaction data record indicating questions presented by a customer, and
an un-structured agent transaction record from a plurality of stored un-structured agent transaction records, the identified un-structured agent transaction record indicating interactions of an agent with a training or customer service interface;
receiving, with the processor, a natural language input from the agent using a human machine interface;
accessing, with the processor, a stored heuristic algorithm;
executing the stored heuristic algorithm, with the processor, to determine:
a current context classifying the customer into a particular category based upon a product type, the heuristic algorithm determining the current context based at least in part on the identified website transaction data record, and
a training resource for providing additional training to the agent, the stored heuristic algorithm determining the training resource based at least in part, on the identified un-structured agent transaction record, the natural language input, and the current context;
determining, with the processor and based on the agent being trained with the training resource, an indication of modified agent behavior; and
training, with the processor, the stored heuristic algorithm using a determined correlation between the training resource and the indication of modified agent behavior.
15. The non-transitory computer readable medium of claim 14 , wherein the identified un-structured agent transaction record comprises a natural language data set.
16. The non-transitory computer readable medium of claim 14 , the acts further comprising calculating the determined correlation as a correlation score based at least in part on the training resource and the indication of modified agent behavior.
17. The non-transitory computer readable medium of claim 14 , wherein the identified un-structured agent transaction record comprises past transactions related to at least one agent.
18. The non-transitory computer readable medium of claim 14 , wherein the identified un-structured agent transaction record comprises a plurality of sales records and the natural language input from the agent comprises a question regarding a pending sales transaction.
19. The non-transitory computer readable medium of claim 14 , wherein the plurality of un-structured agent transaction records are stored in a first memory of an external transaction server and the stored heuristic algorithm is stored in a second memory of an external heuristic server.
20. The non-transitory computer readable medium of claim 14 , wherein:
the plurality of un-structured transaction records are associated with a plurality of agents,
the identified un-structured transaction record comprises real-time interaction data, past interaction data, and customer service data associated with the agent, and
the questions presented by the customer were input via a user terminal, chat window, or natural language input device in communication with a repository storing the plurality of un-structured website transaction data records.