Granted Patent B2

US 12,641,042 · App. 18/545,656 · Granted May 26, 2026

Inventors: Angad Singh Bagga (Punjab, IN); Mayur Ashish Shiradhonkar (Maharashtra, IN); Rajendran Subramanian (Tamil Nadu, IN); Sudhanshu Shukla (Uttas Pradesh, IN); Gopal Krushna Padhi (Odisha, IN); Myra D'Souza (Leander, TX)

Assignee: Fidelity Information Services, LLC

H04L51/02G06F11/0793G06F40/35

Abstract

Claims (99)

1 . A method performed by one or more computers, the method comprising:

joining, by a chatbot, a group conversation in a messaging system in response to receiving an invitation to the group conversation;

receiving, by the chatbot, a natural language request from a first user in the group conversation, the natural language request associated with a current incident;

identifying, by the chatbot, authorization information associated with each user in the group conversation, the authorization information indicating whether each user is eligible to participate in the request;

determining, by the chatbot, one or more authorization requirements associated with each user in the group conversation;

in response to determining that the identified authorization information associated with each user in the group conversation satisfies the determined one or more authorization requirements associated with the natural language request confirming each user in the group conversation is eligible to participate in the request, determining, by the chatbot, a category of the natural language request;

determining, by the chatbot, a processing approach based on the category of the natural language request; and

applying, by the chatbot, the processing approach to generate a response.

2 . The method of claim 1 , wherein determining the category of the natural language request comprises:

applying a classification machine learning (ML) model to the natural language request to select the category from predetermined categories comprising a rule-based category, a generic category, and a knowledge specific category,

and wherein:

the processing approach comprises a rule-based natural language processing algorithm when the natural language request belongs to the rule-based category;

the processing approach comprises a customized ML model when the natural language request belongs to the knowledge specific category; and

the processing approach comprises a fine-tuned large language model (LLM) when the natural language request belongs to the generic category.

3 . The method of claim 2 , wherein:

the natural language request comprises a request to extract information associated with the current incident from the group conversation; and

applying the processing approach comprises applying the rule-based natural language processing algorithm using regular expressions to extract the information from the group conversation.

4 . The method of claim 2 , wherein:

the natural language request comprises a request to find incidents in at least one internal database similar to the current incident; and

applying the processing approach comprises:

applying the customized ML model to determine similarity values between existing incidents stored in the at least one internal database and the current incident; and

selecting one or more of the existing incidents based on the similarity values.

5 . The method of claim 2 , wherein:

the natural language request comprises a request to find a defect that causes the current incident; and

applying the processing approach comprises one or more of:

applying the customized ML model to use regular expressions to find a first defect stored in at least one internal database and associated with an identity of the current incident; or

applying the customized ML model to find a second defect stored in the at least one internal database based on a similarity value between description of the second defect and the current incident.

6 . The method of claim 2 , wherein:

the natural language request comprises a request to find a development change that causes the current incident; and

applying the processing approach comprises:

extracting historical development changes within a predetermined time window from at least one internal database;

applying the customized ML model to determine similarity values between the historical development changes and the current incident; and

selecting one or more of the historical development changes based on the similarity values.

7 . The method of claim 2 , wherein:

the natural language request comprises a request to summarize an existing incident stored in at least one internal database; and

applying the processing approach comprises applying the fine-tuned LLM model to generate a summary of the existing incident.

8 . The method of claim 1 , wherein the authorization information associated with each user in the group conversation comprises one or more of:

whether the user works for an eligible company;

whether the user is a member of an eligible internal team of the eligible company; or

whether the user has authority to conduct an action comprised in the natural language request.

9 . The method of claim 1 , further comprising:

transmitting the response to the group conversation; and

upon request by the first user, sending the response to an email address provided by the first user.

10 . The method of claim 1 , further comprising:

receiving feedback from the first user; and

using the feedback as training data to reinforce one or more machine learning (ML) models comprised in the processing approach.

11 . A system comprising:

one or more computers; and

one or more storage devices storing instructions that when executed by the one or more computers cause the one or more computers to perform operations comprising:

joining, by a chatbot, a group conversation in a messaging system in response to receiving an invitation to the group conversation;

receiving, by the chatbot, a natural language request from a first user in the group conversation, the natural language request associated with a current incident;

identifying, by the chatbot, authorization information associated with each user in the group conversation, the authorization information indicating whether each user is eligible to participate in the request;

determining, by the chatbot, one or more authorization requirements associated with each user in the group conversation;

in response to determining that the identified authorization information associated with each user in the group conversation satisfies the determined one or more authorization requirements associated with the natural language request confirming each user in the group conversation is eligible to participate in the request, determining, by the chatbot, a category of the natural language request;

determining, by the chatbot, a processing approach based on the category of the natural language request; and

applying, by the chatbot, the processing approach to generate a response.

12 . The system of claim 11 , wherein determining the category of the natural language request comprises:

applying a classification machine learning (ML) model to the natural language request to select the category from predetermined categories comprising a rule-based category, a generic category, and a knowledge specific category,

and wherein:

the processing approach comprises a rule-based natural language processing algorithm when the natural language request belongs to the rule-based category;

the processing approach comprises a customized ML model when the natural language request belongs to the knowledge specific category; and

the processing approach comprises a fine-tuned large language model (LLM) when the natural language request belongs to the generic category.

13 . The system of claim 12 wherein:

the natural language request comprises a request to extract information associated with the current incident from the group conversation; and

applying the processing approach comprises applying the rule-based natural language processing algorithm using regular expressions to extract the information from the group conversation.

14 . The system of claim 12 , wherein:

the natural language request comprises a request to find incidents in at least one internal database similar to the current incident; and

applying the processing approach comprises:

applying the customized ML model to determine similarity values between existing incidents stored in the at least one internal database and the current incident; and

selecting one or more of the existing incidents based on the similarity values.

15 . The system of claim 12 , wherein:

the natural language request comprises a request to find a defect that causes the current incident; and

applying the processing approach comprises one or more of:

applying the customized ML model to use regular expressions to find a first defect stored in at least one internal database and associated with an identity of the current incident; or

applying the customized ML model to find a second defect stored in the at least one internal database based on a similarity value between description of the second defect and the current incident.

16 . The system of claim 12 , wherein:

the natural language request comprises a request to find a development change that causes the current incident; and

applying the processing approach comprises:

extracting historical development changes within a predetermined time window from at least one internal database;

applying the customized ML model to determine similarity values between the historical development changes and the current incident; and

selecting one or more of the historical development changes based on the similarity values.

17 . The system of claim 12 , wherein:

the natural language request comprises a request to summarize an existing incident stored in at least one internal database; and

applying the processing approach comprises applying the fine-tuned LLM model to generate a summary of the existing incident.

18 . The system of claim 11 , wherein the authorization information associated with each user in the group conversation comprises one or more of:

whether the user works for an eligible company;

whether the user is a member of an eligible internal team of the eligible company; or

whether the user has authority to conduct an action comprised in the natural language request.

19 . The system of claim 11 , further comprising:

transmitting the response to the group conversation; and

upon request by the first user, sending the response to an email address provided by the first user.

20 . One or more non-transitory computer-readable storage media storing instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:

joining, by a chatbot, a group conversation in a messaging system in response to receiving an invitation to the group conversation;

receiving, by the chatbot, a natural language request from a first user in the group conversation, the natural language request associated with a current incident;

identifying, by the chatbot, authorization information associated with each user in the group conversation, the authorization information indicating whether each user is eligible to participate in the request;

determining, by the chatbot, one or more authorization requirements associated with each user in the group conversation;

in response to determining that the identified authorization information associated with each user in the group conversation satisfies the determined one or more authorization requirements associated with the natural language request confirming each user in the group conversation is eligible to participate in the request, determining, by the chatbot, a category of the natural language request;

determining, by the chatbot, a processing approach based on the category of the natural language request; and

applying, by the chatbot, the processing approach to generate a response.

Assignments (1)

Continuity (1)

References Cited (17)