IP Library › Granted Patent US 12712972

Granted Patent B2

US 12712972 · App. 18/628,307 · Granted Aug 18, 2026

System and method for generating and managing a workflow using webhook technology

Inventors: Shailendra Singh (Thane West, IN); Krishna Rangarao Mamadapur (Pune, IN)

Assignee: Bank of America Corporation

H04M3/5232 G06Q30/016 G10L15/063 G10L15/22 G10L25/18

View Patent ↗

Loading inventors, assignments & file history…

Monitor This Case

Get email alerts when status or documents change.

Order Certified Copies

Most orders are placed with the USPTO same day — all within 24 business hours.

Order via The Patent Place →

Pre-filled with this patent's details

Quick Facts

Patent No.: US 12712972
App. No.: 18/628,307
Granted: Aug 18, 2026
Kind: B2

External Links

USPTO Patent Center↗ Full Patent PDF↗

Abstract

In response to detecting that a voice call has been placed by a first user to an interaction entity, a voice spectrogram is generated of a voice interaction between the first user and a second user associated with the interaction entity and a plurality of signal attributes are extracted from the voice spectrogram. Additionally, a content of the voice interaction is extracted from the voice call. Based on the signal attributes and the content of the voice call, an issue is determined. A resolution entity that is configured to resolve the issue is identified and a webhook payload including an indication of the identified issue is transmitted to the resolution entity for resolution of the issue.

Claims (73)

1 . A system comprising:

a memory configured to store voice calls; and

a processor communicatively coupled to the memory and configured to:

detect that a voice call to an interaction node has been placed by a first user using a voice channel, wherein the voice call comprises voice interaction between the first user and a second user associated with the interaction node;

generate a voice spectrogram of the voice interaction between the first user and the second user, wherein the voice spectrogram is a representation of a voice signal associated with the voice call;

extract a plurality of signal attributes from the voice spectrogram, wherein each signal attribute represents a characteristic of the voice signal;

extract a content of the voice interaction from the voice call, wherein the content of the voice interaction comprises words and sentences spoken by the first user and the second user as part of the voice interaction;

determine, based on the signal attributes and the content, an issue associated with the first user that needs to be resolved, wherein determining the issue associated with the first user comprises:

assign a numerical score to each of the plurality of signal attributes extracted from the voice spectrogram; and

determine that the issue associated with the first user exists in response to the numerical scores assigned to at least one of the signal attributes exceeding a respective threshold;

identify a resolution entity that is configured to resolve the issue;

generate a webhook payload comprising at least an indication of the determined issue associated with the first user; and

route the webhook payload to a pre-configured network address of the resolution entity.

2 . The system of claim 1 , wherein:

a webhook call is pre-configured for each of a plurality of resolution entities; and

the processor is configured to:

generate and transmit a particular webhook payload to a particular resolution entity in response to detecting a respective issue the particular resolution entity is configured to resolve.

3 . The system of claim 1 , wherein the processor is further configured to:

identify a plurality of issues based on the voice call; and

for each of the identified issues, transmit a webhook payload to a respective resolution entity that is configured to resolve the identified issue.

4 . The system of claim 1 , wherein the processor is configured to use an artificial intelligence (AI) algorithm to determine the issue associated with the first user based on the signal attributes and the content.

5 . The system of claim 4 , wherein the processor is further configured to:

receive an indication that the determined issue is incorrect; and

retrain the AI algorithm based on the received indication.

6 . The system of claim 1 , wherein the processor is further configured to generate the voice spectrogram in real-time or near real-time as the voice interaction is being conducted between the first user and the second user at the interaction node.

7 . The system of claim 1 , wherein the plurality of signal attributes extracted from the voice spectrogram comprise one or more of voice modulation, pauses, speech duration, breath, pitch, frequency or loudness.

8 . A method comprising:

detecting that a voice call to an interaction node has been placed by a first user using a voice channel, wherein the voice call comprises voice interaction between the first user and a second user associated with the interaction node;

generating a voice spectrogram of the voice interaction between the first user and the second user, wherein the voice spectrogram is a representation of a voice signal associated with the voice call;

extracting a plurality of signal attributes from the voice spectrogram, wherein each signal attribute represents a characteristic of the voice signal;

extracting a content of the voice interaction from the voice call, wherein the content of the voice interaction comprises words and sentences spoken by the first user and the second user as part of the voice interaction;

determining, based on the signal attributes and the content, an issue associated with the first user that needs to be resolved, wherein determining the issue associated with the first user comprises:

assigning a numerical score to each of the plurality of signal attributes extracted from the voice spectrogram; and

determining that the issue associated with the first user exists in response to the numerical scores assigned to at least one of the signal attributes exceeding a respective threshold;

identifying a resolution entity that is configured to resolve the issue;

generating a webhook payload comprising at least an indication of the determined issue associated with the first user; and

routing the webhook payload to a pre-configured network address of the resolution entity.

9 . The method of claim 8 , wherein:

a webhook call is pre-configured for each of a plurality of resolution entities; and

further comprising:

generating and transmit a particular webhook payload to a particular resolution entity in response to detecting a respective issue the particular resolution entity is configured to resolve.

10 . The method of claim 8 , further comprising:

identifying a plurality of issues based on the voice call; and

for each of the identified issues, transmitting a webhook payload to a respective resolution entity that is configured to resolve the identified issue.

11 . The method of claim 8 , wherein further comprising using an artificial intelligence (AI) algorithm to determine the issue associated with the first user based on the signal attributes and the content.

12 . The method of claim 11 , further comprising:

receiving an indication that the determined issue is incorrect; and

retraining the AI algorithm based on the received indication.

13 . The method of claim 8 , further comprising generating the voice spectrogram in real-time or near real-time as the voice interaction is being conducted between the first user and the second user at the interaction node.

14 . The method of claim 8 , wherein the plurality of signal attributes extracted from the voice spectrogram comprise one or more of voice modulation, pauses, speech duration, breath, pitch, frequency or loudness.

15 . A non-transitory computer-readable medium storing instructions that when executed by a processor causes the processor to: