SYSTEM AND METHOD FOR ADAPTIVE CLOUD CONVERSATION PLATFORM
An adaptive cloud conversation platform capable of making automated decisions regarding when and how to establish ongoing communications with consumers so as to maximize the relationship between the consumer and a given brand. The system has a connection management services layer which determines what communications should be established and how they should be established, an initiation management services layer which determines when communications should be established, and a user management services layer which stores information about consumers and brands for determination of when and how communications should be established. Certain of these services have machine learning algorithms incorporated into them trained to perform analyses of the particular type of operation handled by that service. The outputs of each service can be used as inputs to other services, such that a network of machine learnings algorithms is created which determines when and how to establish ongoing communications with consumers.
1 . An adaptive cloud conversation platform, comprising:
a computing device comprising a memory, a processor, and a non-volatile data storage device;
a consumer profile comprising at least some of either behaviors of the consumer or preferences of the consumer, or at least a combination of both;
a conversation manager comprising a first plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive a determination that a callback should be made to a consumer;
retrieve a consumer profile for the consumer;
process the plurality of preferences through a first machine learning algorithm to select a channel through which to conduct the callback with the consumer; and
forward the channel selection to a schedule manager;
a schedule manager comprising a second plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive the channel selection from the conversation manager;
retrieve the consumer profile;
process the plurality of preferences through a second machine learning algorithm to select a time at which to conduct the callback with the consumer through the selected channel; and
schedule a callback with the consumer at the selected time through the selected channel.
2 . The platform of claim 1 , further comprising:
a callback manager comprising a third plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive a callback schedule; and
execute a callback at a selected time through a selected channel as indicated in the callback schedule.
3 . The platform of claim 1 , further comprising:
a multi-channel dynamic interaction engine comprising a fourth plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
collect context and user data regarding user callbacks and outcomes;
collect heterogeneous sensor data from a modular array of sensors that may include GPS or location sensors, physical activity sensors, or other sensor data, if available;
process the context, user, and sensor data, through a machine learning algorithm to determine an optimal callback channel or medium, and to determine or improve callback timing for a user; and
forward optimal callback channel or medium data, and callback timing data, to a conversation manager.
4 . The platform of claim 1 , further comprising a consumer context manager comprising a third plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive text from a conversation with a consumer;
retrieve a consumer profile for the consumer comprising at least some of either behaviors of the consumer or preferences of the consumer, or at least a combination of both;
process the text and the plurality of behaviors of the consumer through a third machine learning algorithm to determine whether a callback to the consumer should be made; and
where the determination is that a callback should be made, forward the determination to the conversation manager as the determination that a callback should be made to a consumer.
5 . The platform of claim 4 , further comprising a session manager comprising a fourth plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive the text;
process the text through a fourth machine learning algorithm to determine a consumer sentiment; and
forward the determined sentiment to the consumer context manager as an additional input to the third machine learning algorithm's determination as to whether a callback to the consumer should be made.
6 . The platform of claim 4 , wherein:
the conversation manager is further configured to:
process the text through a fifth machine learning algorithm to determine a consumer goal, need, or intent; and
forward the determined goal, need, or intent to the consumer context manager as an additional input to the third machine learning algorithm's determination as to whether a callback to the consumer should be made.
7 . The platform of claim 4 , further comprising:
an event rules database stored on the non-volatile data storage device, the event rules database comprising rules for triggering communications with consumers based on events occurring outside of a conversation; and
an event manager comprising a fifth plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive notification of an event;
match the event to a rule in the event rules database; and
forward the matched event to the conversation manager as the determination that a callback should be made to the consumer.
8 . The platform of claim 7 , wherein:
the event manager is further configured to:
retrieve the consumer profile for the consumer from the consumer profile database;
process the consumer profile and one or more rules from the event rules database through a sixth machine learning algorithm to determine a new rule for triggering communications with the consumer; and
store the new rule in the event rules database.
9 . The platform of claim 4 , further comprising:
a brand environment database stored on the non-volatile data storage device, the brand environment database comprising brand information related to conversations with consumers of the brand;
an environment manager comprising a sixth plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
retrieve a consumer profile for a consumer from the consumer profile database;
retrieve the brand information from the brand environment database;
process the consumer profile and the brand information through a seventh machine learning algorithm to determine whether a campaign of communications should be established with a plurality of consumers; and
forward the determination to the conversation manager as the determination that a callback should be made to the consumer.
10 . The platform of claim 9 , further comprising:
a consumer manager comprising a seventh plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
retrieve the consumer profile for the consumer from the consumer profile database;
receive a fitness parameter from the environment manager;
process the consumer profile and the fitness parameter through an eighth machine learning algorithm to identify opportunities for proactive conversations with the consumer; and
forward the identified opportunities to the consumer context manager as an additional input to the first machine learning algorithm's selection of the channel through which the callback should be made.
11 . A method for operating an adaptive cloud conversation platform, comprising the steps of:
using a conversation manager operating on a computing device comprising a memory, a processor, and a non-volatile data storage device to:
receive a determination that a callback should be made to a consumer;
retrieve a consumer profile for the consumer comprising at least some of either behaviors of the consumer or preferences of the consumer, or at least a combination of both;
process the plurality of preferences through a first machine learning algorithm to select a channel through which to conduct the callback with the consumer; and
forward the channel selection to a schedule manager;
using a schedule manager operating on the computing device to:
receive the channel selection from the conversation manager;
retrieve the consumer profile;
process the plurality of preferences through a second machine learning algorithm to select a time at which to conduct the callback with the consumer through the selected channel; and
schedule a callback with the consumer at the selected time through the selected channel.
12 . The method of claim 11 , further comprising the steps of using a callback manager to:
receive a callback schedule; and
execute a callback at a selected time through a selected channel as indicated in the callback schedule.
13 . The method of claim 11 , further comprising the steps of using a multi-channel dynamic interaction engine to:
collect context and user data regarding user callbacks and outcomes;
collect heterogeneous sensor data from a modular array of sensors that may include GPS or location sensors, physical activity sensors, or other sensor data, if available;
process the context, user, and sensor data, through a machine learning algorithm to determine an optimal callback channel or medium, and to determine or improve callback timing for a user; and
forward optimal callback channel or medium data, and callback timing data, to a conversation manager.
14 . The method of claim 11 , further comprising the steps of using a consumer context manager operating on the computing device to:
receive text from a conversation with a consumer;
retrieve a consumer profile for the consumer comprising at least some of either behaviors of the consumer or preferences of the consumer, or at least a combination of both;
process the text and the plurality of behaviors of the consumer through a third machine learning algorithm to determine whether a callback to the consumer should be made; and
where the determination is that a callback should be made, forward the determination to the conversation manager as the determination that a callback should be made to a consumer.
15 . The method of claim 14 , further comprising the steps of using a session manager operating on the computing device to:
receive the text;
process the text through a fourth machine learning algorithm to determine a consumer sentiment; and
forward the determined sentiment to the consumer context manager as an additional input to the third machine learning algorithm's determination as to whether a callback to the consumer should be made.
16 . The method of claim 14 , further comprising the steps of configuring the conversation manager to:
process the text through a fifth machine learning algorithm to determine a consumer goal, need, or intent; and
forward the determined goal, need, or intent to the consumer context manager as an additional input to the third machine learning algorithm's determination as to whether a callback to the consumer should be made.
17 . The method of claim 14 , further comprising the steps of:
using an event manager operating on the computing device to:
receive notification of an event;
match the event to a rule in the event rules database stored on the non-volatile data storage device, the event rules database comprising rules for triggering communications with consumers based on events occurring outside of a conversation; and
forward the matched event to the conversation manager as the determination that a callback should be made to the consumer.
18 . The method of claim 17 , further comprising the steps of configuring the event manager to:
retrieve the consumer profile for the consumer from the consumer profile database;
process the consumer profile and one or more rules from the event rules database through a sixth machine learning algorithm to determine a new rule for triggering communications with the consumer; and
store the new rule in the event rules database.
19 . The method of claim 14 , further comprising the steps of:
using an environment manager operating on the computing device to:
retrieve a consumer profile for a consumer from the consumer profile database;
retrieve brand information from a brand environment database stored on the non-volatile data storage device, the brand environment database comprising brand information related to conversations with consumers of the brand;
process the consumer profile and the brand information through a seventh machine learning algorithm to determine whether a campaign of communications should be established with a plurality of consumers; and
forward the determination to the conversation manager as the determination that a callback should be made to the consumer.
20 . The method of claim 19 , further comprising the steps of:
using a consumer manager operating on the computing device to:
retrieve the consumer profile for the consumer from the consumer profile database;
receive a fitness parameter from the environment manager;
process the consumer profile and the fitness parameter through an eighth machine learning algorithm to identify opportunities for proactive conversations with the consumer; and
forward the identified opportunities to the consumer context manager as an additional input to the first machine learning algorithm's selection of the channel through which the callback should be made.