IP Library › Granted Patent US 12671663
Granted Patent B2
US 12671663 · App. 18/648,777 · Granted Jun 30, 2026

Conversational commerce optimization for AI-based virtual assistants

Inventors: Satish Gunjal (Pune, IN); Manish Pasrija (Faridabad, IN); Anoop Kumar Gopinatha (Issaquah, WA); Himanshu Srivastava (Gurugram, IN); Elodie Perrigaud (Nanterre, FR)
Assignee: Accenture Global Solutions Limited
H04L51/02G06F16/2425
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 12671663
App. No.
18/648,777
Granted
Jun 30, 2026
Kind
B2
Abstract

Implementations of the present disclosure provide receiving, by a conversational commerce optimizer system, a user input from a user, the user input being provided to a chatbot of an e-commerce platform, providing a refined query based on the user input, selecting, by the conversational commerce optimizer system, a foundation model that is to be queried using the refined query, determining a set of search results from a database that stores data representative of products available through the e-commerce platform, prompting the foundation model using a prompt that is generated based on the refined query and the set of search results, the foundation model providing an output responsive to the prompt, and providing a response to the user through the user interface of the chatbot, the response being generated based on the output.

Claims (52)

1 . A computer-implemented method for facilitating interactions between a user and a chatbot of an electronic commerce (e-commerce) platform, the method comprising:

receiving, by a query refiner of a conversational commerce optimizer system, a first user input from a user, the first user input being provided during a chat session through a user interface of the chatbot of the e-commerce platform;

providing, by the query refiner, a first refined query based on the first user input;

selecting, by a Generative Artificial Intelligence (GAI) selection module of the conversational commerce optimizer system, a first foundation model that is to be queried using the first refined query;

determining, by the query refiner, a first set of search results from a database that stores data representative of products available through the e-commerce platform;

prompting the first foundation model using a first prompt that is generated based on the first refined query and the first set of search results, the first foundation model providing a first output responsive to the first prompt;

providing a first response to the user through the user interface of the chatbot, the first response being generated based on the first output;

receiving, by the query refiner, a second user input from the user;

providing, by the query refiner, a second refined query based on the second user input;

selecting, by the GAI selection module, a second foundation model that is to be queried using the second refined query;

prompting the second foundation model using a second prompt that is generated based on the second refined query, the second foundation model providing a second output responsive to the second prompt; and

providing a second response to the user though the user interface of the chatbot, the second response being generated based on the second output.

2 . The computer-implemented method of claim 1 , wherein the first response is provided by validating the first output and formatting the first output.

3 . The computer-implemented method of claim 1 , wherein the first set of search results is generated by comparing a set of user input embeddings to embeddings stored in the database.

4 . The computer-implemented method of claim 1 , wherein the first set of search results is provided from an initial set of search results determined based on the first refined query.

5 . The computer-implemented method of claim 1 , herein the second refined query is provided at least partially based on the first user input.

6 . The computer-implemented method of claim 1 , wherein the second prompt is provided at least partially based on the first user input and the first output.

7 . A system, comprising:

one or more processors; and

a computer-readable storage device coupled to the one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for facilitating interactions between a user and a chatbot of an electronic commerce (e-commerce) platform, the operations comprising:

receiving, by a query refiner of a conversational commerce optimizer system, a first user input from a user, the first user input being provided during a chat session through a user interface of the chatbot of the e-commerce platform;

providing, by the query refiner, a first refined query based on the first user input;

selecting, by a Generative Artificial Intelligence (GAI) selection module of the conversational commerce optimizer system, a first foundation model that is to be queried using the first refined query;

determining, by the query refiner, a first set of search results from a database that stores data representative of products available through the e-commerce platform;

prompting the first foundation model using a first prompt that is generated based on the first refined query and the first set of search results, the first foundation model providing a first output responsive to the first prompt;

providing a first response to the user through the user interface of the chatbot, the first response being generated based on the first output;

receiving, by the query refiner, a second user input from the user;

providing, by the query refiner, a second refined query based on the second user input;

selecting, by the GAI selection module, a second foundation model that is to be queried using the second refined query;

prompting the second foundation model using a second prompt that is generated based on the second refined query, the second foundation model providing a second output responsive to the second prompt; and

providing a second response to the user though the user interface of the chatbot, the second response being generated based on the second output.

8 . The system of claim 7 , wherein the first response is provided by validating the first output and formatting the first output.

9 . The system of claim 7 , wherein the first set of search results is generated by comparing a set of user input embeddings to embeddings stored in the database.

10 . The system of claim 7 , wherein the first set of search results is provided from an initial set of search results determined based on the first refined query.

11 . The system of claim 7 , wherein the second refined query is provided at least partially based on the first user input.

12 . The system of claim 7 , wherein the second prompt is provided at least partially based on the first user input and the first output.

13 . Computer-readable storage media coupled to the one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for facilitating interactions between a user and a chatbot of an electronic commerce (e-commerce) platform, the operations comprising:

receiving, by a query refiner of a conversational commerce optimizer system, a first user input from a user, the first user input being provided during a chat session through a user interface of the chatbot of the e-commerce platform;

providing, by the query refiner, a first refined query based on the first user input;

selecting, by a Generative Artificial Intelligence (GAI) selection module of the conversational commerce optimizer system, a first foundation model that is to be queried using the first refined query;

determining, by the query refiner, a first set of search results from a database that stores data representative of products available through the e-commerce platform;

prompting the first foundation model using a first prompt that is generated based on the first refined query and the first set of search results, the first foundation model providing a first output responsive to the first prompt;

providing a first response to the user through the user interface of the chatbot, the first response being generated based on the first output;

receiving, by the query refiner, a second user input from the user;

providing, by the query refiner, a second refined query based on the second user input;

selecting, by the GAI selection module, a second foundation model that is to be queried using the second refined query;

prompting the second foundation model using a second prompt that is generated based on the second refined query, the second foundation model providing a second output responsive to the second prompt; and

providing a second response to the user though the user interface of the chatbot, the second response being generated based on the second output.

14 . The computer-readable storage media of claim 13 , wherein the first response is provided by validating the first output and formatting the first output.

15 . The computer-readable storage media of claim 13 , wherein the first set of search results is generated by comparing a set of user input embeddings to embeddings stored in the database.

16 . The computer-readable storage media of claim 13 , wherein the first set of search results is provided from an initial set of search results determined based on the first refined query.

17 . The computer-readable storage media of claim 13 , wherein the second refined query is provided at least partially based on the first user input.