US 12,450,433 B2
Real time key conversational metrics prediction and notability
Bharat S. Raj, Kerala (IN); Eric Jee-Keng Dunn, Milpitas, CA (US); Dmytro Kovalchuk, Mill Valley, CA (US); Brenton William D'Adamo, Tampa, FL (US); and Benjamin S. Shan, San Francisco, CA (US)
Assigned to Sutherland Global Services Inc., Pittsford, NY (US)
Filed by Sutherland Global Services Inc., Pittsford, NY (US)
Filed on Feb. 21, 2023, as Appl. No. 18/172,001.
Application 18/172,001 is a continuation in part of application No. 16/863,613, filed on Apr. 30, 2020, granted, now 11,587,552, issued on Feb. 21, 2023.
Claims priority of provisional application 62/841,140, filed on Apr. 30, 2019.
Prior Publication US 2023/0297780 A1, Sep. 21, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 40/284 (2020.01); G06F 40/216 (2020.01); G06Q 30/02 (2023.01)
CPC G06F 40/284 (2020.01) [G06F 40/216 (2020.01); G06Q 30/0281 (2013.01)] 52 Claims
OG exemplary drawing
 
1. A method for alerting a manager device to an occurrence of an event an agent device during a conversation between the agent device and an external party, the method comprising:
receiving transcript data during a conversation between the agent device and the external party;
normalizing the transcript data, wherein normalizing the transcript data comprises anonymizing personal data within the transcript data;
inputting the normalized transcript data into a machine learning model, the machine learning model trained to identify an inflection point in the conversation based on training samples, the training samples including labels indicating whether an inflection point has occurred with reference to textual transcript data and auxiliary transcript data of each given training sample, wherein the training samples comprise normalized message text encoded to vector embeddings that have auxiliary data other than the message text concatenated to the vector embeddings, and wherein the auxiliary data comprises time-based data;
receiving, as output from the machine learning model, a measure of notability of the normalized transcript data;
determining whether the measure of notability corresponds to an inflection point; and
responsive to determining that the measure of notability corresponds to an inflection point, alerting the manager device.