SYSTEMS AND METHODS FOR SUPERVISING AND IMPROVING GENERATIVE AI CONTENT

1 . A system comprising:

a memory configured to store instructions for an artificial intelligence (AI) system; and

a processor coupled to the memory and configured to read the instructions from the memory to cause the system to perform operations, the operations comprising:

receiving an AI generated content at a neural network;

determining, using the neural network and at least one prompt, a concept in the AI generated content;

receiving source content for the concept in the AI generated content;

identifying, using the neural network, the source content, the concept, and a list of error types, an error in the AI generated content; and

inserting the AI generated content with the error into a queue, wherein the AI generated content in the queue is displayed on an editable AI interface.

2 . The system of claim 1 , wherein the operations further comprise:

displaying the AI generated content on the editable AI interface;

receiving, via the editable AI interface, instructions to modify the AI generated content based on the error; and

providing the modified AI generated content for display on another user device.

3 . The system of claim 2 , wherein the operations further comprise:

providing the modified AI generated content to at least one neural network in a plurality of neural networks for retraining the at least one neural network using the modified AI generated content.

4 . The system of claim 1 , wherein the operations for identifying the error in the AI generated content further comprise:

querying a database for the source content that includes the concept; and

comparing a concept in the source content with the concept in the AI generated content to identify the error.

5 . The system of claim 1 , wherein the neural network is a language model.

6 . The system of claim 1 , wherein the operations further comprise:

identifying one or more trigger words within the AI generated content, wherein the one or more trigger words are in a pre-defined list.

7 . The system of claim 1 , wherein operations further comprise:

receiving an updated AI generated content that corresponds to the AI generated content with the error;

identifying that there is no error in the updated AI generated content; and

removing the AI generated content from the queue, wherein the updated AI generated content is generated in response to new source content available in a database.

8 . A method comprising:

determining, using a neural network executing on a processor and at least one prompt, a concept in an AI generated content;

querying a database for source content from which the AI generated content was generated, wherein the source content is associated with the concept;

identifying, using the neural network, the source content, and the concept, an error in the AI generated content; and

inserting the AI generated content with the error into a queue, wherein the AI generated content in the queue is displayed on an editable AI interface.

9 . The method of claim 8 , further comprising:

displaying the AI generated content on an editable AI interface; and

modifying, using the editable AI interface, the AI generated content based on the error.

10 . The method of claim 9 , further comprising:

providing the modified AI generated content to the neural network for retraining the neural network using the modified AI generated content.

11 . The method of claim 8 , further comprising:

providing, to the neural network, a list of errors in a configuration file, wherein the list of errors includes the error.

12 . The method of claim 8 , further comprising:

identifying one or more trigger words within the AI generated content, wherein the one or more trigger words are in a pre-defined list; and

wherein the determining is based on identifying the one or more trigger words within the AI generated content.

13 . The method of claim 8 , further comprising:

receiving an updated AI generated content;

identifying that there is no error in the updated AI generated content; and

removing the AI generated content from the queue, wherein the updated AI generated content is generated in response to the database being updated with a new source content.

14 . The method of claim 8 , wherein the AI generated content comprises at least one of:

a patient summary;

a discharge barrier; or

a physician note.

15 . A non-transitory computer-readable medium having instructions thereon, that when executed by a processor, cause the processor to perform operations comprising:

receiving an AI generated summary;

determining, using a neural network and at least one prompt, a concept in the AI generated summary;

querying a database for source content for the concept in the AI generated summary;

identifying, using the neural network, the source content, and the AI generated summary, an error in the AI generated summary; and

inserting the AI generated summary in a queue available in an editable AI interface.

16 . The non-transitory computer-readable medium of claim 15 , wherein the operations further comprising:

displaying the AI generated summary on the editable AI interface; and

receiving instructions to modify, using an editable AI interface, the AI generated summary based on the error.

17 . The non-transitory computer-readable medium of claim 16 , wherein the operations further comprising:

providing the modified AI generated summary to the neural network for retraining the neural network using the modified AI generated summary.

18 . The non-transitory computer-readable medium of claim 17 , wherein the AI generated summary is parsed by a language model.

19 . The non-transitory computer-readable medium of claim 15 , the operations further comprising:

identifying one or more trigger words within the AI generated summary, wherein the one or more trigger words are stored in a configuration file; and

identifying the error in the AI generated summary when the AI generated summary includes the one or more trigger words.

20 . The non-transitory computer-readable medium of claim 15 , the operations further comprising:

receiving an updated AI generated summary;

identifying that there is no error in the updated AI generated summary; and

removing the AI generated summary from the queue, wherein the updated AI generated summary is generated in response to new source content available in the database.