Responses to requests for information

1. A computer-implemented method, comprising:

training a machine learning model to predict when documents will be available from webpages, thereby providing a trained machine learning model;

receiving, by a server, a first request for first information and a second request for second information different from the first information;

determining, by the server, that a first document including the first information has not been released;

determining, by the server, that a second document different from the first document and including the second information has not been released;

generating, by the trained machine learning model, predictions of when the first document will be available from a first webpage and when the second document will be available from a second webpage, the first webpage being different from the second webpage;

configuring, based on the predictions, a web crawler included on the server to:

access the first webpage at a first frequency based on one of the predictions; and

access the second webpage at a second frequency different from the first frequency and based on another of the predictions;

accessing, by the web crawler, the first webpage at the first frequency, and thereby obtaining, by the web crawler, the first document from the first webpage after the first document is released;

accessing, by the web crawler, the second webpage at the second frequency and thereby obtaining, by the web crawler, the second document from the second webpage after the second document is released;

retaining the first request until the first information is released and retaining the second request until the second information is released;

parsing, by the server, the first document, when released, to obtain the first information from the first document;

based on the parsing the first document, generating, by the server, a first response to the first request, the first response including the first information;

parsing, by the server, the second document, when released, to obtain the second information from the second document; and

based on the parsing the second document, generating, by the server, a second response to the second request, the second response including the second information.

2. The method of claim 1 , wherein the first response includes a computer-generated text message or a computer-generated audio message.

3. The method of claim 1 , further comprising:

determining when the first document is scheduled to be released; and

selecting a storage location in a computer memory for storing the first request based on a length of time until the first document is released.

4. The method of claim 1 , further comprising generating a message indicating when the first response will be provided.

5. The method of claim 4 , further comprising determining when the first response will be provided based on a scheduled release of the first document.

6. The method of claim 1 , further comprising:

generating a possible request, the possible request corresponding to the first request; and

providing the possible request for display at a graphical interface,

wherein the receiving the first request includes receiving a selection, via the graphical interface, of the possible request.

7. The method of claim 6 , wherein the generating the possible request includes:

processing, with the trained machine learning model, data relating to a user; and

based on the processing, using the trained machine learning model to output the possible request.

8. The method of claim 7 , wherein the data includes a prior request for information associated with the user.

9. The method of claim 7 , wherein the data includes a purchase history associated with the user.

10. The method of claim 7 , further comprising:

determining, with the trained machine learning model and based on the possible request, that the first document will include the first information.

11. The method of claim 6 ,

wherein the generating the possible request includes receiving, via another graphical interface, a selection of an unreleased document, the selection of the unreleased document corresponding to the first document.

12. The method of claim 1 , wherein the first response includes an indication that the first document has been released.

13. The method of claim 1 , wherein the parsing, by the server, the first document includes applying natural language understanding to text of the first document.

14. The method of claim 1 , wherein the first response includes a recommendation, based on the first request and the first information, to perform or not to perform a financial action.

15. A system, comprising:

one or more processors; and

non-transitory computer-readable instructions that, when executed by the one or more processors, cause the system to:

train a machine learning model to predict when documents will be available from webpages, thereby providing a trained machine learning model;

receive, at a first time, a first request for first information;

receive, at a second time, a second request for second information;

determine, using a web crawler, that a first document containing the first information has not been released at the first time and, based thereon, retain the first request until a third time that is after the first time;

determine, using the web crawler, that a second document different from the first document and containing the second information has not been released at the second time;

generate, by the trained machine learning model, predictions of when the first document will be available from a first webpage and when the second document will be available from a second webpage, the first webpage being different from the second webpage;

configure, based on the predictions, the web crawler to:

access the first webpage at a first frequency based on one of the predictions; and

access the second webpage at a second frequency different from the first frequency and based on another of the predictions;

access, by the web crawler, the first webpage at the first frequency, and thereby obtain, by the web crawler, the first document from the first webpage after the first document is released at a fourth time that is after the first time;

access, by the web crawler, the second webpage at the second frequency and thereby obtain, by the web crawler, the second document from the second webpage after the second document is released;

parse text of the first document using natural language understanding to obtain the first information from the first document;

generate, at or before the third time, a first response to the first request, the first response including the first information and being triggered by the first information being obtained from the first document;

parse text of the second document using natural language understanding to obtain the second information from the second document; and

generate a second response to the second request, the second response including the second information and being triggered by the second information being obtained from the second document.

16. The system of claim 15 , wherein the first response includes a computer-generated text message or a computer-generated audio message.

17. The system of claim 16 , including further instructions which, when executed by the one or more processors, cause the system to:

generate another message indicating when the first response will be provided based on a scheduled release of the first document.

18. The system of claim 17 , wherein the first response includes an indication that the first document has been released.

19. The system of claim 15 , wherein the first response includes a recommendation, based on the first request and the first information, to perform or not to perform a financial action.

20. A computer-implemented method, comprising:

training a machine learning model to predict when documents will be available from webpages, thereby providing a trained machine learning model;

receiving, by a server at a first time, a first request for first information associated with a financial action;

receiving, by the server at a second time, a second request for second information;

determining, using a web crawler, that a first document including the first information has not been released at the first time;

determining, using the web crawler, that a second document different from the first document and containing the second information has not been released at the second time;

determining, by the server, that the first document is scheduled to be released at a third time that is after the first time;

retaining the first request until a fourth time that is after the first time and at or after the third time, including:

selecting a storage location in a computer memory for storing the first request based on a length of time between the first time and the third time; and

storing the first request in the storage location;

generating, by the trained machine learning model, predictions of when the first document will be available from a first webpage and when the second document will be available from a second webpage, the first webpage being different from the second webpage;

configuring, by the server and based on the predictions, the web crawler to:

access the first webpage at a first frequency based on one of the predictions; and

access the second webpage at a second frequency different from the first frequency and based on another of the predictions;

accessing, by the web crawler, the first webpage at the first frequency, and thereby obtaining, by the web crawler, the first document from the first webpage after the first document is released and at or after the third time;

accessing, by the web crawler, the second webpage at the second frequency and thereby obtaining, by the web crawler, the second document from the second webpage after the second document is released;

parsing, by the server using natural language understanding, text of the first document to obtain the first information from the first document;

based on the parsing, generating by the server, at or after the third time, a first response to the first request, the first response including a text message or an audio message including:

the first information; and

a recommendation to perform or not to perform the financial action;

parsing, by the server, text of the second document using natural language understanding to obtain the second information from the second document; and

generating, by the server a second response to the second request, the second response including the second information and being triggered by the second information being obtained from the second document.