Granted Patent B2

US 12,342,423 · App. 17/861,163 · Granted Jun 24, 2025

Inventors: Phi Hoang Nguyen (Bellevue, WA); Veeraseshamamba Venigalla (Southlake, TX)

Assignee: T-Mobile USA, Inc.

H04W8/20H04W8/24H04W8/26H04W16/22

Abstract

Claims (75)

1. A non-transitory computer-readable medium carrying instructions that, when executed by a computing system, cause the computing system to perform operations to modify a condition or state of mobile devices, the operations comprising:

receiving a set of interactions between an agent of a telecommunications service provider and a subscriber of the telecommunications service provider,

wherein the set of interactions comprises one or more of: messages exchanged between the agent and the subscriber, or transcribed audio of an exchange between the agent and the subscriber, and

wherein the set of interactions relates to use, by the subscriber, of a mobile device to access a network provided by the telecommunications service provider;

identifying and accessing subscriber data for the subscriber based, at least in part, on the set of interactions between the agent and the subscriber,

wherein the subscriber data includes a coverage map indicating coverage of the network in multiple locations;

determining a usage pattern characterizing the use of the mobile device to access the network provided by the telecommunications service provider,

wherein the usage pattern relates to locations where the mobile device accesses the network and activities conducted via the network; and

using the set of interactions, the subscriber data, and the usage pattern, applying a trained network usage analysis model to recommend at least one action to respond to a condition or state associated with the mobile device,

wherein the at least one action includes modifying a device state of the mobile device, installing an update on the mobile device, or recommending a different mobile device to access the network, and

wherein the network usage analysis model comprises a machine learning model that has been trained using training data comprising multiple sets of interactions,

wherein each set of interactions corresponds to a corresponding subscriber, a corresponding mobile device, and a corresponding agent of the telecommunications service provider,

wherein each set of interactions comprise one or more of: messages exchanged between the corresponding agent and the corresponding subscriber, or transcribed audio of an exchange between the corresponding agent and the corresponding subscriber,

wherein each set of interactions relate to use, by the corresponding subscriber, of the corresponding mobile device to access a network provided by the telecommunications service provider, and

wherein each set of interactions is associated with a corresponding action taken in response to the set of interactions, the corresponding action relating to a condition or state of the corresponding mobile device.

2. The non-transitory computer-readable medium of claim 1 , wherein the operations further comprise:

applying intent analysis to the received set of interactions,

wherein the subscriber data is identified based in part on the intent analysis, and

wherein the trained network usage analysis model further uses the intent analysis to recommend the at least one action.

3. The non-transitory computer-readable medium of claim 1 , wherein at least some of the operations are performed while the interactions between the agent and the subscriber are occurring.

4. The non-transitory computer-readable medium of claim 1 , wherein the subscriber data includes demographic information about the subscriber or financial information about the subscriber.

5. The non-transitory computer-readable medium of claim 1 , wherein the operations further comprise:

causing display of the subscriber data, the recommendation of the at least one action, or both the subscriber data and the recommendation of the at least one action.

6. The non-transitory computer-readable medium of claim 1 , wherein the condition or state associated with the mobile device relates to the network provided by the telecommunications service provider or a component or node of the network.

7. The non-transitory computer-readable medium of claim 1 , wherein the recommended at least one action includes modifying a component or node of the network provided by the telecommunications service provider, replacing a component or node of the network provided by the telecommunications service provider, or adding a microcell, a picocell, or a femtocell to the network provided by the telecommunications service provider.

8. The non-transitory computer-readable medium of claim 1 , wherein the operations further comprise:

identifying and accessing technical data for the mobile device, the network provided by the telecommunications service provider, or both the mobile device and the network.

9. A computer-implemented method of modifying a condition or state of mobile devices, the method comprising:

receiving a set of interactions between an agent of a telecommunications service provider and a subscriber of the telecommunications service provider,

wherein the set of interactions comprises one or more of: messages exchanged between the agent and the subscriber, or transcribed audio of an exchange between the agent and the subscriber, and

wherein the set of interactions relates to use, by the subscriber, of a mobile device to access a network provided by the telecommunications service provider;

identifying and accessing subscriber data for the subscriber based, at least in part, on the set of interactions between the agent and the subscriber,

wherein the subscriber data includes a coverage map indicating coverage of the network in multiple locations;

determining a usage pattern characterizing the use of the mobile device to access the network provided by the telecommunications service provider,

wherein the usage pattern relates to locations where the mobile device accesses the network and activities conducted via the network; and

using the set of interactions, the subscriber data, and the usage pattern, applying a network usage analysis model to recommend at least one action to respond to a condition or state associated with the mobile device,

wherein the at least one action includes modifying a device state of the mobile device, installing an update on the mobile device, or recommending a different mobile device to access the network, and

wherein the network usage analysis model comprises a machine learning model that has been trained using training data comprising multiple sets of interactions,

wherein each set of interactions corresponds to a corresponding subscriber, a corresponding mobile device and a corresponding agent of the telecommunications service provider,

wherein each set of interactions comprise one or more of: messages exchanged between the corresponding agent and the corresponding subscriber, or transcribed audio of an exchange between the corresponding agent and the corresponding subscriber,

wherein each set of interactions relate to use, by the corresponding subscriber, of the corresponding mobile device to access a network provided by the telecommunications service provider, and

wherein each set of interactions is associated with a corresponding action taken in response to the set of interactions, the corresponding action relating to a condition or state of the corresponding mobile device.

10. The computer-implemented method of claim 9 , further comprising:

applying intent analysis to the received set of interactions,

wherein the subscriber data is identified based in part on the intent analysis, and

wherein the applied network usage analysis model further uses the intent analysis to recommend the at least one action.

11. The computer-implemented method of claim 9 , wherein at least some of the method is performed while the interactions between the agent and the subscriber are occurring.

12. The computer-implemented method of claim 9 , wherein the subscriber data includes demographic information about the subscriber or financial information about the subscriber.

13. The computer-implemented method of claim 9 , further comprising:

causing display of the subscriber data, the recommendation of the at least one action, or both the subscriber data and the recommendation of the at least one action.

14. The computer-implemented method of claim 9 , wherein the condition or state associated with the mobile device relates to the network provided by the telecommunications service provider or a component or node of the network.

15. The computer-implemented method of claim 9 , wherein the recommended at least one action includes modifying a component or node of the network provided by the telecommunications service provider, replacing a component or node of the network provided by the telecommunications service provider, or adding a microcell, a picocell, or a femtocell to the network provided by the telecommunications service provider.

16. The computer-implemented method of claim 9 , further comprising:

identifying and accessing technical data for the mobile device, the network provided by the telecommunications service provider, or both the mobile device and the network.

17. A non-transitory computer-readable medium carrying instructions that, when executed by a computing system, cause the computing system to perform operations to train a machine learning model to generate recommendations to modify a condition or state associated with mobile devices, the operations comprising:

receiving multiple sets of interactions between agents of a telecommunications service provider and subscribers of the telecommunications service provider,

wherein each set of interactions of the multiple sets of interactions comprise one or more of: messages exchanged between an agent and a subscriber, or transcribed audio of an exchange between the agent and the subscriber,

wherein each set of interactions of the multiple sets of interactions relate to use, by the subscriber, of a mobile device to access a network provided by the telecommunications service provider, and

wherein each set of interactions of the multiple sets of interactions are associated with a corresponding action taken in response to the set of interactions, the corresponding action relating to a condition or state of the mobile device;

receiving supplemental data for at least some sets of interactions of the multiple sets of interactions, the supplemental data comprising at least one of:

subscriber data comprising a coverage map indicating coverage of the network in multiple locations, demographic data, or device data, or

a usage pattern characterizing the use of the mobile device to access the network provided by the telecommunications service provider;

using the multiple sets of interactions and the supplemental data to generate a training dataset,

wherein intent analysis is applied to each of the multiple sets of interactions to determine an intent of each set of interactions;

using the training dataset to train a machine learning model to generate recommended actions to respond to conditions or states associated with mobile devices,

applying the trained machine learning model to recommend at least one action to respond to a particular condition or state associated with a particular mobile device,

wherein the recommended at least one action is based on a particular set of interactions between a particular subscriber and a particular agent related to use, by the particular subscriber, of the particular mobile device to access the network provided by the telecommunications service provider.

18. The non-transitory computer-readable medium of claim 17 , wherein the machine learning model is further trained to:

identify and access data sources containing the supplemental data; and

identify the conditions or states associated with the mobile devices.

19. The non-transitory computer-readable medium of claim 17 , wherein the operations further comprise:

evaluating an accuracy of the trained machine learning model using a testing dataset,

wherein the testing dataset includes a plurality of sets of interactions between subscribers of the telecommunications service provider and agents of the telecommunications service provider; and

retraining the machine learning model with the accuracy does not exceed a threshold accuracy,

wherein retraining the machine learning model includes at least one of: training the machine learning model at least a second time using the generated training dataset, resampling at least a portion of the generated training dataset, or training the machine learning model using a different dataset.

Assignments (1)

Continuity (1)

References Cited (70)