Method for order taking using interactive virtual human agents

• US 8,054,951 B1
• Filed: 02/10/2006
• Issued: 11/08/2011
• Est. Priority Date: 04/29/2005
• Status: Expired due to Fees

• Current Assignee: Ignite Media Solutions LLC (Ignite-Tek LLC)
• Original Assignee: Ignite Media Solutions LLC (Ignite-Tek LLC)
• Inventors: Thomas, Anthony, Winslow, Douglas
• Primary Examiner(s): Tsang; Fan
• Assistant Examiner(s): King; Simon
• Application Number:
  US11/350,895
• Time in Patent Office: 2,097 Days
• Field of Search: 379/69, 379/88.16, 379/88.25, 379/88.01, 379/88.18, 704/251, 704/275
• US Class Current: 379/88.18
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions
• CPC Class Codes: H04M 3/4936 Speech interaction details speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/5166 in combination with interactive voice response systems or voice portals, e.g. as front-ends

Abstract

Methods combine speech recognition with telephonic sales order processing, in order to automate and improve the efficiency of telephone sales orders, by responding appropriately to vocalizations, pausing for interruption, repeating responses when appropriate, automatically transferring calls to human agents or different automated systems at appropriate points in a transaction, and collecting information. A method accepts incoming sales order calls and appropriately handles order processing. A method connects outgoing calls to customers who have disconnected before completing sales order calls, and appropriately handles sales and order processing.

Claims

1. A method for telephonic sales order processing, the method steps comprising:

establishing a call connection with a called party;

playing a recorded greeting over the call connection;

monitoring a vocal response of a called party;

repeating the recorded greeting, if the monitored vocal response does not match a user-defined expected quantity;

identifying the meaning of the vocal response of the called party substituting responses with a default response if the responses can not be verified;

playing a recording in response to the vocal response of the called party;

repeating the recording if an audible interruption occurs before the playing of the recording is completed;

and

storing the vocal response on a non-transitory computer-readable medium using a computer.

2. The method of claim 1, wherein a response of the called party includes a Dual-Tone-Mutli-Frequency (DTMF) response.

3. The method of claim 1, wherein any input not DTMF and less than 1 second long qualifies as a monitored vocal response that does not match the user-defined expected quantity.

4. The method of claim 1, wherein a user-defined option allows the vocal response less than 1 second long to qualify as not matching the user-defined expected quantity.

5. The method of claim 1, wherein if the vocal response cannot be verified above a user-defined confidence level, a further recording asks the called party to repeat themselves.

6. The method of claim 1, wherein barge-in during the playing of a recording can be ignored based on configuration.

7. The method of claim 1, wherein during the playing of a recording only DTMF tones are recognized based on user configuration.

8. The method of claim 1, wherein several recordings are played in sequence.

9. The method of claim 8, wherein the vocal response ignored while any one of the several recordings is played, but wherein after the any one of the several recordings is complete no subsequent recording is played until an additional vocal response is detected.

10. The method of claim 9, wherein if the detected additional vocal response matches a user-defined expected quantity, the next recording in the sequence is played.

11. The method of claim 8, wherein barge-in during the playing of a recording can be ignored based on configuration.

12. The method of claim 1, wherein if the vocal response of the called party cannot be identified after a user-defined number of attempts, the call connection is transferred to a live agent.

13. The method of claim 1, wherein several recordings are played in sequence without waiting for a response.

14. The method of claim 13, wherein if an audible interruption is detected for longer than 1 second during the playing of a recording, an alternative recording is played and then the recording that was playing at the time of the audible interruption is re-played.

15. A method for telephonic sales order processing, the method steps comprising:

connecting an outgoing call wherein no information is known about a called party except a call destination;

playing a recorded greeting during the call to the called party;

monitoring vocal responses of the called party;

identifying the meaning of the vocal responses of the called party substituting responses with a default response if the responses can not be verified;

playing a recording in response to the vocal responses of the called party;

and

storing the vocal responses on a non-transitory computer-readable medium using a computer.

16. The method of claim 15, wherein several recordings are played in sequence without any interruption.

17. The method of claim 16, wherein one of said vocal responses is detected but ignored while a recording is played, but wherein after the recording is complete no subsequent recording is played until a subsequent vocal response is detected.

18. The method of claim 17, wherein if the detected subsequent vocal response matches a user-defined expected quantity, the next recording in the sequence is played.

19. The method of claim 17, wherein if the detected subsequent vocal response does not match a user-defined expected quantity, the recording is repeated.

20. The method of claim 15, wherein if the detected subsequent vocal response of the called party cannot be identified after a user-defined number of attempts, the call is transferred to a live agent.

21. The method of claim 15, wherein a request by the called party not to be called is recognized and a telephone number associated with the called party is added to a do not call list.

22. A method for telephonic sales order processing, the method steps comprising:

making a call connection;

playing a recorded greeting over the call connection;

monitoring vocal responses of a called party;

identifying the meaning of the vocal responses of the called party;

substituting the vocal responses of the called party with an user-defined default response if the vocal responses cannot be verified above a user-defined confidence level;

playing a recording in response to the vocal responses or user-defined default response of the called party;

and

storing the vocal responses or an indication of the user-defined default response on a non-transitory computer-readable medium using a computer.