Systems and methods for transferring messaging to automation
The present disclosure relates generally to facilitating routing of communications. More specifically, techniques are provided to dynamically transfer messaging between a network device and a terminal device to a type of bot based on intents identified from the messaging. Further, techniques are provided to track performance of the selected type of bot during automation.
1. A computer-implemented method, comprising:
receiving a set of messages in real-time during a conversation between a network device and a terminal device, wherein the terminal device is associated with a live agent, and wherein the set of messages are received as the set of messages are exchanged;
determining an intent associated with the conversation, wherein the intent is determined in real-time as the set of messages are exchanged;
using the intent and the set of messages as input to a trained machine learning model to determine a score corresponding to the conversation, wherein the score represents a sentiment associated with the network device, and wherein the trained machine learning model is trained using a dataset of input messages and corresponding scores representing different sentiments;
determining that the score corresponding to the conversation exceeds a threshold value, wherein the threshold value corresponds to an allocation of conversations to different bots or live agents;
dynamically transferring the conversation from the live agent to a bot, wherein the conversation is dynamically transferred as a result of the score exceeding the threshold value;
receiving feedback with regard to the conversation, wherein the feedback corresponds to the set of messages exchanged between the network device and the terminal device and to new messages exchanged between the network device and the bot; and
training a new machine learning model to dynamically determine future intents associated with future conversations, wherein the new machine learning model is trained using the conversation, the sentiment, and the feedback.
2. The computer-implemented method of claim 1 , wherein the feedback is received through textual or non-textual attributes associated with the set of messages exchanged during the conversation.
3. The computer-implemented method of claim 1 , wherein the conversation is dynamically transferred from the live agent to the bot automatically without requiring live agent intervention.
4. The computer-implemented method of claim 1 , wherein the trained machine learning model identifies from the set of messages an anchor associated with a polarity, and wherein the polarity is used by the trained machine learning model to determine the score.
5. The computer-implemented method of claim 1 , wherein dynamically transferring the conversation from the live agent to the bot further comprises:
updating an interface displayed on the terminal device to present an option for transferring the conversation to the bot, wherein the option is presented with a confidence score corresponding to a likelihood of the bot being able to address the intent; and
detecting selection of the option through the interface, wherein when the selection is detected, the conversation is transferred to the bot.
6. The computer-implemented method of claim 1 , wherein the new machine learning model is further trained to correlate historical messages with different intents based on the feedback.
7. The computer-implemented method of claim 1 , further comprising:
obtaining agent feedback that is reflective of a performance of the bot in identifying and addressing the intent; and
using the agent feedback to further train the new machine learning model.
8. A system, comprising:
one or more processors; and
memory storing thereon instructions that, as a result of being executed by the one or more processors, cause the system to:
receive a set of messages in real-time during a conversation between a network device and a bot, terminal device, wherein the terminal device is associated with a live agent, and wherein the set of messages are received as the set of messages are exchanged;
determine an intent associated with the conversation, wherein the intent is determined in real-time as the set of messages are exchanged;
use the intent and the set of messages as input to a trained machine learning model to determine a score corresponding to the conversation, wherein the score represents a sentiment associated with the network device, and wherein the trained machine learning model is trained using a dataset of input messages and corresponding scores representing different sentiments;
determine that the score corresponding to the conversation exceeds a threshold value, wherein the threshold value corresponds to an allocation of conversations to different bots or live agents;
dynamically transfer the conversation from the live agent to a bot, wherein the conversation is dynamically transferred as a result of the score exceeding the threshold value;
receive feedback with regard to the conversation, wherein the feedback corresponds to the set of messages exchanged between the network device and the terminal device and to new messages exchanged between the network device and the bot; and
train a new machine learning model to dynamically determine future intents associated with future conversations, wherein the new machine learning model is trained using the conversation, set of messages, the sentiment, and the feedback.
9. The system of claim 8 , wherein the feedback is received through textual or non-textual attributes associated with the set of messages exchanged during the conversation.
10. The system of claim 8 , wherein the conversation is dynamically transferred from the live agent to the bot automatically without requiring live agent intervention.
11. The system of claim 8 , wherein the trained machine learning model identifies from the set of messages an anchor associated with a polarity, and wherein the polarity is used by the trained machine learning model to determine the score.
12. The system of claim 8 , wherein the instructions that cause the system to dynamically transfer the conversation from the live agent to the bot further cause the system to:
update an interface displayed on the terminal device to present an option for transferring the conversation to the bot, wherein the option is presented with a confidence score corresponding to a likelihood of the bot being able to address the intent; and
detect selection of the option through the interface, wherein when the selection is detected, the conversation is transferred to the bot.
13. The system of claim 8 , wherein the new machine learning model is further trained to correlate historical messages with different intents based on the feedback.
14. The system of claim 8 , wherein the instructions further cause the system to:
obtain agent feedback that is reflective of a performance of the bot in identifying and addressing the intent; and
use the agent feedback to further train the new machine learning model.
15. A non-transitory, computer-readable storage medium storing thereon executable instructions that, as a result of being executed by one or more processors of a computer system, cause the computer system to:
receive a set of messages in real-time during a conversation between a network device and a bot, terminal device, wherein the terminal device is associated with a live agent, and wherein the set of messages are received as the set of messages are exchanged;
determine an intent associated with the conversation, wherein the intent is determined in real-time as the set of messages are exchanged;
use the intent and the set of messages as input to a trained machine learning model to determine a score corresponding to the conversation, wherein the score represents a sentiment associated with the network device, and wherein the trained machine learning model is trained using a dataset of input messages and corresponding scores representing different sentiments;
determine that the score corresponding to the conversation exceeds a threshold value, wherein the threshold value corresponds to an allocation of conversations to different bots or live agents;
dynamically transfer the conversation from the live agent to a bot, wherein the conversation is dynamically transferred as a result of the score exceeding the threshold value;
receive feedback with regard to the conversation, wherein the feedback corresponds to the set of messages exchanged between the network device and the terminal device and to new messages exchanged between the network device and the bot; and
train a new machine learning model to dynamically determine future intents associated with future conversations, wherein the new machine learning model is trained using the conversation, set of messages, the sentiment, and the feedback.
16. The non-transitory, computer-readable storage medium of claim 15 , wherein the feedback is received through textual or non-textual attributes associated with the set of messages exchanged during the conversation.
17. The non-transitory, computer-readable storage medium of claim 15 , wherein the conversation is dynamically transferred from the live agent to the bot automatically without requiring live agent intervention.
18. The non-transitory, computer-readable storage medium of claim 15 , wherein the trained machine learning model identifies from the set of messages an anchor associated with a polarity, and wherein the polarity is used by the trained machine learning model to determine the score.
19. The non-transitory, computer-readable storage medium of claim 15 , wherein the executable instructions that cause the computer system to dynamically transfer the conversation from the live agent to the bot further cause the computer system to:
update an interface displayed on the terminal device to present an option for transferring the conversation to the bot, wherein the option is presented with a confidence score corresponding to a likelihood of the bot being able to address the intent; and
detect selection of the option through the interface, wherein when the selection is detected, the conversation is transferred to the bot.
20. The non-transitory, computer-readable storage medium of claim 15 , wherein the new machine learning model is further trained to correlate historical messages with different intents based on the feedback.
21. The non-transitory, computer-readable storage medium of claim 15 , wherein the executable instructions further cause the computer system to:
obtain agent feedback that is reflective of a performance of the bot in identifying and addressing the intent; and
use the agent feedback to further train the new machine learning model.