SYSTEMS AND METHODS FOR USING CLIENT SYSTEM INTELLIGENCE FOR DISTRIBUTED SERVICE SYSTEM CONFIGURATION

1 . A method for executing a service of a server computer system for a client system, comprising:

defining, at the server computer system, a client system descriptor file that includes a set of signals suitable for input into a large language machine learning model (LLM);

collecting, by the server computer system, a first set of data associated with a first client system from one or more source systems, wherein the first set of data is representative of the set of signals for the client system;

compressing, by the server computer system, the first set of data into the set of signals for a first client system descriptor file allocated for the client system;

executing, by the server computer system, a query using at least the first client system descriptor file as a portion of input into the LLM, the query associated with a service of the server computer system provided to the client system; and

in response to obtaining an output by the LLM, executing at least an action by the service of the server computer system.

2 . The method of claim 1 , wherein the collecting further comprises:

accessing first data associated with the client system from a local data store of a server computer system;

querying one or more remote data stores for a second data associated with the client system; and

wherein the first data and the second data comprise the first set of data.

3 . The method of claim 1 , wherein the collecting further comprises:

performing a second machine learning model analysis of a second set of data associated with the client system, wherein a result of the analysis comprises a higher order vector or matrix representing a second output of a second machine learning model;

joining the higher order vector or matrix to the first set of data prior to execution of the query.

4 . The method of claim 1 , wherein compressing the first set of data into the set of signals further comprises:

transforming, by the server computer system, one or more values of the first set of data into a binned value for a first signal in the set of signals.

5 . The method of claim 1 , wherein the query is a retrieval augmented generation (RAG) query, the server computer system accesses a set of second client system descriptor files, and wherein the set of second client system descriptor files forms another portion of the input into the LLM for execution of the query.

6 . The method of claim 1 , wherein the first client system descriptor file input into the LLM for execution of the query is cached, and wherein a second query is executed by the server computer system using the LLM using the cached input of the first client system descriptor file.

7 . The method of claim 1 , wherein executing at least the action by the service of the server computer system comprises:

adjusting an operational characteristic of the service system based on the output by the LLM;

generating and transmitting an alert message to an operator responsible for the service system;

generating a graphical user interface that displays the answer output by the LLM to an operator associated with the service system.

8 . The method of claim 1 , wherein prior to execution of the query by the LLM, the LLM is trained in a first unsupervised training phase to respond to natural language queries, and the method further comprises:

performing a second training phase of the LLM based on annotated training data supplied by the server computer, the second training phase being a supervised training phase using annotated data comprising query responses generated within an operational domain of the server computer system and training sets of signals corresponding to the client system descriptor file.

9 . The method of claim 8 , further comprising:

determining, by the server computer system, at least one of the set of signals in the client descriptor file that provides a contribution to a plurality of answers generated by the LLM below a minimum threshold value;

defining, at the server computer system, a new client system descriptor file that includes a new set of signals without the at least one of the set of signals, wherein a total number of signals in the new client system descriptor file is less than a total number of signals in the client system descriptor file; and

initiating the second training phase of the LLM using the signal set defined by the new client system descriptor file.

10 . The method of claim 8 , wherein the second training phase is periodically performed to refine the LLM over time.

11 . A non-transitory computer readable storage medium storing instructions, which when executed by a server computer system, causes the server computer system to perform operations for executing a service of the server computer system for a client system, the operations comprising:

defining, at the server computer system, a client system descriptor file that includes a set of signals suitable for input into a large language machine learning model (LLM);

collecting, by the server computer system, a first set of data associated with a first client system from one or more source systems, wherein the first set of data is representative of the set of signals for the client system;

compressing, by the server computer system, the first set of data into the set of signals for a first client system descriptor file allocated for the client system;

executing, by the server computer system, a query using at least the first client system descriptor file as a portion of input into the LLM, the query associated with a service of the server computer system provided to the client system; and

in response to obtaining an output by the LLM, executing at least an action by the service of the server computer system.

12 . The non-transitory computer readable storage medium of claim 11 , wherein the operations for collecting further comprise:

performing a second machine learning model analysis of a second set of data associated with the client system, wherein a result of the analysis comprises a higher order vector or matrix representing a second output of a second machine learning model;

joining the higher order vector or matrix to the first set of data prior to execution of the query.

13 . The non-transitory computer readable storage medium of claim 11 , wherein the query is a retrieval augmented generation (RAG) query, the server computer system accesses a set of second client system descriptor files, and wherein the set of second client system descriptor files forms another portion of the input into the LLM for execution of the query.

14 . The non-transitory computer readable storage medium of claim 11 , wherein the first client system descriptor file input into the LLM for execution of the query is cached, and wherein a second query is executed by the server computer system using the LLM using the cached input of the first client system descriptor file.

15 . The non-transitory computer readable storage medium of claim 11 , wherein prior to execution of the query by the LLM, the LLM is trained in a first unsupervised training phase to respond to natural language queries, and the operations further comprise:

performing a second training phase of the LLM based on annotated training data supplied by the server computer, the second training phase being a supervised training phase using annotated data comprising: query responses generated within an operational domain of the server computer system and training sets of signals corresponding to the client system descriptor file.

16 . A system for executing a service of a server computer system for a client system, the system comprising:

a memory; and

one or more processors coupled with the memory configured to perform operations, comprising:

defining, at the server computer system, a client system descriptor file that includes a set of signals suitable for input into a large language machine learning model (LLM);

collecting, by the server computer system, a first set of data associated with a first client system from one or more source systems, wherein the first set of data is representative of the set of signals for the client system;

compressing, by the server computer system, the first set of data into the set of signals for a first client system descriptor file allocated for the client system;

executing, by the server computer system, a query using at least the first client system descriptor file as a portion of input into the LLM, the query associated with a service of the server computer system provided to the client system; and

in response to obtaining an output by the LLM, execute at least an action by the service of the server computer system.

17 . The system of claim 16 , wherein the one or more processors are further configured to perform operations, comprising:

performing a second machine learning model analysis of a second set of data associated with the client system, wherein a result of the analysis comprises a higher order vector or matrix representing a second output of a second machine learning model;

joining the higher order vector or matrix to the first set of data prior to execution of the query.

18 . The system of claim 16 , wherein the query is a retrieval augmented generation (RAG) query, the server computer system accesses a set of second client system descriptor files, and wherein the set of second client system descriptor files forms another portion of the input into the LLM for execution of the query.

19 . The system of claim 16 , wherein the first client system descriptor file input into the LLM for execution of the query is cached, and wherein a second query is executed by the server computer system using the LLM using the cached input of the first client system descriptor file.

20 . The system of claim 16 , wherein prior to execution of the query by the LLM, the LLM is trained in a first unsupervised training phase to respond to natural language queries, and the one or more processors are further configured to perform operations, comprising:

performing a second training phase of the LLM based on annotated training data supplied by the server computer, the second training phase being a supervised training phase using annotated data comprising: query responses generated within an operational domain of the server computer system and training sets of signals corresponding to the client system descriptor file.