Claims processing with extracted claim variables

1 . A provider computing system comprising:

a communication interface structured to communicatively couple the provider computing system to a network;

a claims database storing claim statements for a plurality of claims, the claim statements comprising a plurality of claim variables;

an orchestration circuit storing computer-executable instructions embodying one or more machine learning models;

at least one processor; and

at least one memory, the at least one memory storing instructions that, when executed by the at least one processor, cause the at least one processor to:

receive a first claim statement comprising natural language text corresponding to a new claim;

cause, by the orchestration circuit, the one or more machine learning models to parse the natural language text in the first claim statement to extract one or more claim variables from the first claim statement;

after determining that at least one claim variable of a predetermined set of claim variables is missing from the one or more claim variables, generate, by the orchestration circuit, a response comprising a request for a second claim statement requesting information regarding the at least one missing claim variable; and

transmit the request for a second claim statement.

2 . The provider computing system of claim 1 , wherein the instructions further cause the at least one processor to train, by the orchestration circuit, the one or more machine learning models based on a subset of the claim statements such that the one or more machine learning models outputs a predicted plurality of claim variables based on the claim statements.

3 . The provider computing system of claim 2 , wherein the instructions further cause the at least one processor to generate, by a synthetic statement generating circuit of the provider computing system, a new plurality of synthetic claim statements.

4 . The provider computing system of claim 3 , wherein the instructions further cause the at least one processor to train, by the orchestration circuit, the one or more machine learning models based on the plurality of synthetic claim statements such that the one or more machine learning models outputs a predicted plurality of claim variables based on the synthetic claim statement.

5 . The provider computing system of claim 1 , wherein the instructions further cause the at least one processor to:

provide the one or more claim variables to a claims processing server;

receive, from the claims processing server, a claim outcome; and

cause a user device to display the claim outcome.

6 . The provider computing system of claim 5 , wherein the instructions further cause the at least one processor to:

receive the second claim statement comprising additional natural language text;

cause, by the orchestration circuit, the one or more machine learning models to parse the additional natural language text in the second claim statement to extract one or more additional claim variables; and

provide the one or more additional claim variables to the claims processing server.

7 . A method comprising:

communicatively coupling, by a communication interface, a provider computing system to a network;

receiving, from a user device, a first claim statement comprising natural language text corresponding to a first claim;

parsing, by one or more machine learning models associated with an orchestration circuit of the provider computing system, the natural language text in the first claim statement to extract one or more claim variables from the first claim statement;

comparing the one or more claim variables to a predetermined set of required claim variables;

responsive to determining that at least one required claim variable of the set of required claim variables is not present in the one or more claim variables, generating, by the orchestration circuit of the provider computing system, a request for a second claim statement;

transmitting the request for the second claim statement to the user device;

receiving the second claim statement comprising additional natural language text corresponding to the first claim;

parsing, by the one or more machine learning models associated with the orchestration circuit of the provider computing system, the additional natural language text in the second claim statement to extract one or more additional claim variables from the second claim statement;

comparing the one or more claim variables and the one or more additional claim variables to the predetermined set of required claim variables; and

responsive to determining that each of the required claim variables of the set of required claim variables are in at least one of the one or more claim variables or the one or more additional claim variables, causing the user device to display one or more claim variables and the one or more additional claim variables.

8 . The method of claim 7 , further comprising training, by the orchestration circuit, the one or more machine learning models based on at least the first claim statement such that the one or more machine learning models outputs a predicted plurality of claim variables based on the first claim statement.

9 . The method of claim 8 , further comprising generating, by a synthetic statement generating circuit of the provider computing system, a new plurality of synthetic claim statements.

10 . The method of claim 9 , further comprising training, by the orchestration circuit, the one or more machine learning models based on a subset of the synthetic claim statements such that the one or more machine learning models outputs a predicted plurality of claim variables based on the synthetic claim statements.

11 . The method of claim 7 , further comprising:

providing the one or more claim variables to a claims processing server;

receiving, from the claims processing server, a claim outcome; and

causing the user device to display the claim outcome.

12 . The method of claim 7 , further comprising:

generating at least one graphical feature displaying a question regarding the at least one required claim variable that is not present in the one or more claim variables; and

causing the user device to display the at least one graphical feature.

13 . A computing system for processing insurance claims comprising: a communication interface structured to communicatively couple the computing system to a network;

an orchestration circuit storing computer-executable instructions embodying one or more machine learning models;

at least one processor; and

at least one memory, the at least one memory storing instructions that, when executed by the at least one processor, cause the at least one processor to:

receive a first claim statement comprising natural language text corresponding to a new claim;

cause, by the orchestration circuit, the one or more machine learning models to parse the natural language text in the first claim statement to extract one or more claim variables;

provide the one or more claim variables to a claims processing server via the network;

receive, from the claims processing server, at least one of: an indication that at least one claim variable is missing from the one or more claim variables, or an indication that the one or more claim variables is complete;

responsive to receiving the indication that the at least one claim variable is missing from the one or more claim variables, generate, by the orchestration circuit, a response comprising a request for a second claim statement;

transmit the request for a second claim statement;

responsive to receiving the indication that the one or more claim variables is complete, provide the one or more claim variables to a claims processing server;

receive, from the claims processing server, a claim outcome; and

cause the user device to display the claim outcome.

14 . The computing system of claim 13 , wherein the instructions further cause the at least one processor to train, by the orchestration circuit, the one or more machine learning models based on at least one of the first claim statement or the second claim statement such that the one or more machine learning models outputs a predicted plurality of claim variables based on the first claim statement or the second claim statement.

15 . The computing system of claim 14 , wherein the instructions further cause the at least one processor to:

generate, by a synthetic statement generating circuit of the provider computing system, a new plurality of synthetic claim statements; and

train, by the orchestration circuit, the one or more machine learning models based on the plurality of synthetic claim statements such that the one or more machine learning models outputs a predicted plurality of claim variables based on the synthetic claim statements.

16 . The computing system of claim 13 , wherein generating, by the orchestration circuit, the response comprising the request for the second claim statement comprises generating a first graphical feature displaying a question regarding at least one missing claim variable.

17 . The computing system of claim 16 , wherein generating, by the orchestration circuit, the response comprising the request for the second claim statement comprises generating a second graphical feature displaying a multiple choice input, the multiple choice input including choices regarding the missing claim variable.

18 . The computing system of claim 13 , wherein the instructions further cause the at least one processor to:

receive the second claim statement comprising additional natural language text;

cause, by the orchestration circuit, the one or more machine learning models to parse the additional natural language text in the second claim statement to extract one or more additional claim variables; and

provide the one or more additional claim variables to the claims processing server.

19 . The computing system of claim 13 , wherein the instructions further cause the at least one processor to:

receive, from the claims processing server, an indication that a combination of the one or more claim variables and the one or more additional claim variables is complete; and

provide the one or more claim variables and the one or more additional claim variables to the claims processing server.