Interactive Voice Response Transcoding

Patent metadata

Field	Value
Publication number	US-2025106321-A1
Application number	US-202318474905-A
Country	US
Kind code	A1
Filing date	Sep 26, 2023
Priority date	Sep 26, 2023
Publication date	Mar 27, 2025
Grant date	—

Abstract

Techniques are disclosed that relate to a computer system implementing an interactive voice response (IVR) transcoder. The computer system may receive voice input from an interactive voice response (IVR) system, or other channel. The computer system converts the received input to a request having a common format supported by an artificial intelligence (AI) core including one or more virtual agents. The computer system routes the request to the AI core operable to handle requests specified in the common format. The computer system receives, by the interface module, response data including data (e.g. text data) responsive to the second request. The computer system may implement a conversational artificial intelligence (CAI) platform. The computer system may receive text input from a chat-based input source. The computer system may route another request based on the text input to the AI core, the request being specified in the common text format.

First claim

What is claimed is:

1 . A method, comprising:
receiving, at an interface module of a computer system, first input from a chat-based interface; routing, by the interface module, a first request based on the first input to an artificial intelligence (AI) core operable to handle the first request, the AI core including one or more virtual agents and the first request being specified in a common input format; receiving, by the interface module from the AI core, first response data responsive to the first request, the first response data being specified in a common output format; outputting, by the interface module based on the first response data, first output data to the chat-based interface; receiving, by the interface module, second input from an interactive voice response (IVR) system; converting, by the interface module, the second input to a second request having the common input format; routing, by the interface module, the second request to the AI core, the AI core being operable to handle the second request; receiving, by the interface module, second response data responsive to the second request, the second response data being specified in the common output format; converting, by the interface module, the second response data to second output data compatible with the IVR system; and sending, by the interface module, the second output data to the IVR system.

2 . The method of claim 1 , wherein converting the second input includes:
sending, by the interface module, the second input to a voice-to-text service; receiving, by the interface module from the voice-to-text service, text transcribed by the voice-to-text service from the second input; and including, by the interface module, the transcribed text in the second request.

3 . The method of claim 1 , wherein converting the second response data includes:
generating, by the interface module, a transcript based on second response data; sending, by the interface module, the transcript to a text-to-voice service; receiving, by the interface module from the text-to-voice service, a synthesized speech output; and including, by the interface module, the synthesized speech output in the second output data.

4 . The method of claim 1 , wherein the second output data includes text information usable by the IVR system to generate synthesized sound.

5 . The method of claim 1 , wherein the second input from the IVR includes an indication of one or more key presses made by a telephone device communicating with the IVR.

6 . The method of claim 1 , wherein the AI core is operable to generate the first response data and/or the second response data to include information describing a menu with one or more options for user selection.

7 . The method of claim 6 , further comprising:
receiving, by the IVR system, the second response data; and outputting, by the IVR system based on the second response data, an audio menu prompt such that a user of the IVR system can select a given one of the one or more options using:
a press of a telephone keypad button corresponding to the given option; or
a voice command specifying the given option.

8 . The method of claim 1 , further comprising:
storing, by the computer system, state information for a given request submitted via a particular input source; and wherein the state information is usable by the computer system to transfer the given request from the particular input source to a different input source while the given request is in process.

9 . The method of claim 8 , wherein the state information is usable by the computer system to transfer the given request from the IVR system to the chat-based interface in response to a break in communication with the IVR system.

10 . The method of claim 1 , wherein a first subset of the one or more virtual agents utilize machine-learning techniques, and a second subset of the one or more virtual agents utilize rule-based techniques.

11 . A non-transitory, computer-readable storage medium storing program instructions executable by a computer system to perform operations comprising:
receiving, at an interface module of the computer system, first input from a first input source; routing, by the interface module, a first request based on the first input to an artificial intelligence (AI) core operable to handle the first request, the AI core including one or more virtual agents and the first request being specified in a common input format; receiving, by the interface module from the AI core, first response data responsive to the first request, the first response data being specified in a common output format; outputting, by the interface module based on the first response data, first output data to first input source; receiving, by the interface module, second input from a second input source, where the first and second input source have different input formats; routing, by the interface module based on the second input, a second request to the AI core, the second request being specified in the common input format; receiving, by the interface module, second response data responsive to the second request, the second response data being specified in the common output format; and sending, by the interface module based on the second response data, second output data to the second input source.

12 . The non-transitory, computer-readable storage medium of claim 11 , wherein:
a first input format of the first input source is a chat-based interface that receives text input from a user; and a second input format of the second input source is an interactive voice response (IVR) system that receives voice input and/or telephonic key presses from the user.

13 . The non-transitory, computer-readable storage medium of claim 11 , wherein the first input source is a home assistant device.

14 . The non-transitory, computer-readable storage medium of claim 12 , wherein the operations further comprise:
creating, by the interface module, the second request, wherein the creating includes converting the received second input into the common input format; and creating, by the interface module, the second output data, wherein the creating includes converting the second response data to the common output format.

15 . The non-transitory, computer-readable storage medium of claim 14 , wherein the second input source is an IVR system, and wherein creating the second output data includes:
generating, by the interface module, a transcript based on second response data; sending, by the interface module, the transcript to a text-to-voice service; receiving, by the interface module from the text-to-voice service, a synthesized speech output; and including, by the interface module, the synthesized speech output in the second output data.

16 . The non-transitory, computer-readable storage medium of claim 11 , wherein the operations further comprise:
storing, by the computer system, state information for a given request sent by a particular input source, where the state information is indicative of actions taken by a given virtual agent relative to the given request; and transferring, by the computer system, the given request and the state information from the particular input source to a different input source while the given request is in process.

17 . A system, comprising:
one or more processor circuits; a memory storing program instructions executable by the one or more processor circuits to implement a conversational artificial intelligence (CAI) platform that includes:
an artificial intelligence (

Assignees

• Paypal Inc

Inventors

• Reyes Oliver Jamero
• Manthena Vinaykanth
• Kua James Siang-Jin
• Rooney Niall
• Xi Liang
• Aswal Ajay

Classifications

• H04M3/5166

  in combination with interactive voice response systems or voice portals, e.g. as front-ends · CPC title

• H04M3/493 Primary

  Interactive information services, e.g. directory enquiries {; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals} · CPC title

• G10L15/26

  Speech to text systems (G10L15/08 takes precedence) · CPC title

• G10L13/08

  Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination · CPC title