Recommendation generation with user values
A system can obtain a first prompt as output from inputting an alert about a computer system to a first retrieval-augmented generation system (RAG). The system can obtain a first answer as output from inputting the first prompt to a first large language model (LLM). The system can obtain a value maintained by an entity associated with the computing system as output from inputting the alert to a second RAG. The system can obtain a second answer as output from inputting the first answer, the value, and a second prompt to a second LLM, wherein the second LLM comprises the first LLM or another LLM different from the first LLM. The system can obtain a third answer as output from inputting the second answer, user information associated with the entity, and a third prompt to a third LLM. The system can make the third answer available to the entity.
1 . A system, comprising:
at least one processor; and
at least one memory that stores executable instructions that, when executed by the at least one processor, facilitate performance of operations, comprising:
identifying an alert regarding operation of a computing system other than the system;
obtaining a first prompt as output from inputting the alert to a first retrieval-augmented generation system;
obtaining a first answer as output from inputting the first prompt to a first large language model;
obtaining a value maintained by an entity associated with the computing system as output from inputting the alert to a second retrieval-augmented generation system, wherein the second retrieval-augmented generation system comprises the first retrieval-augmented generation system or another retrieval-augmented generation system different from the first retrieval-augmented generation system;
obtaining a second answer as output from inputting the first answer, the value maintained by the entity associated with the computing system, and a second prompt to a second large language model, wherein the second large language model comprises the first large language model or another large language model different from the first large language model;
obtaining a third answer as output from inputting the second answer, information about the entity that is separate from the value, and a third prompt to a third large language model, wherein the third large language model comprises the second large language model or another large language model different from the second large language model; and
enabling the third answer to be accessible via a device associated with the entity.
2 . The system of claim 1 , wherein the inputting of the alert to the second retrieval-augmented generation system results in output of a group of values that comprise the value, and wherein inputting the first answer, the value maintained by the entity associated with the computing system, and the second prompt to the second large language model comprises:
performing iterations of inputting the first answer, respective values of the values, and the second prompt to respective large language models of a group of at least one large language model that comprises the second large language model to obtain the second answer.
3 . The system of claim 1 , wherein the second retrieval-augmented generation system comprises a keywords vector data store of user values that comprise the value maintained by the entity associated with the computing system.
4 . The system of claim 3 , wherein the operations further comprise:
storing information from unstructured text data in the keywords vector data store.
5 . The system of claim 3 , wherein the operations further comprise:
storing information from structured text data, audio data, image data, video data, or interaction data in the keywords vector data store.
6 . The system of claim 3 , wherein the operations further comprise:
determining the value maintained by the entity associated with the computing system based on implicit information that is derived from actions taken by the entity, wherein the actions taken by the entity are separate from explicitly identifying the value.
7 . The system of claim 3 , wherein the second retrieval-augmented generation system comprises telemetry data that corresponds to the computing system.
8 . The system of claim 1 , wherein the operations further comprise:
receiving answer feedback data that is associated with the entity; and
updating the first retrieval-augmented generation system, the first large language model, the second retrieval-augmented generation system, the value maintained by the entity associated with the computing system, or the second large language model based on the answer feedback data.
9 . A method, comprising:
sending, by a system comprising at least one processor, an alert regarding operation of a computing system other than the system to a first retrieval-augmented generation system to produce a first prompt;
inputting, by the system, the first prompt to a first large language model to produce a first answer;
sending, by the system, the alert to a second retrieval-augmented generation system to produce a value, wherein the value indicates a principle that has merit to an entity that is associated with the computing system, and wherein the second retrieval-augmented generation system comprises the first retrieval-augmented generation system or another retrieval-augmented generation system;
inputting, by the system, the first answer, the value, and a second prompt to a second large language model to produce a second answer, wherein the second large language model comprises the first large language model or another large language model;
inputting, by the system, the second answer, information about the entity that is separate from the value, and a third prompt to a third large language model to produce a third answer, wherein the third large language model comprises the second large language model or another large language model; and
making, by the system, the third answer accessible to a device associated with the entity.
10 . The method of claim 9 , wherein the alert is a first alert, and further comprising:
after the making of the third answer accessible to the device associated with the entity, and based on determining that the value has changed, storing, by the system, an updated value; and
determining, by the system, a fourth answer based on a second alert regarding operation of the computing system based on the updated value.
11 . The method of claim 9 , wherein a group of values comprises the value, and further comprising:
determining, by the system, a ranking of respective values of the group of values; and
wherein the third answer is determined based on the ranking.
12 . The method of claim 9 , further comprising:
identifying, by the second retrieval-augmented generation system, telemetry information that is associated with the computing system; and
determining, by the second retrieval-augmented generation system, the value based on the telemetry information.
13 . The method of claim 9 , wherein the value comprises a value type, and wherein at least one parameter applicable to the second retrieval-augmented generation system is modified based on the value type.
14 . The method of claim 9 , wherein the value comprises a value type, and wherein at least one parameter applicable to the second large language model is modified based on the value type.
15 . A non-transitory computer-readable medium comprising instructions that, in response to execution, cause a system comprising at least one processor to perform operations, comprising:
communicating an alert regarding operation of a computing system other than the system to a first retrieval-augmented generation system resulting in obtaining a first prompt;
communicating the first prompt to a first large language model resulting in obtaining a first answer;
communicating the alert to a second retrieval-augmented generation system resulting in obtaining a value, wherein the value indicates a concept that has been specified to be desirable to an entity that is associated with the computing system, and wherein the second retrieval-augmented generation system comprises the first retrieval-augmented generation system or another retrieval-augmented generation system;
communicating the first answer, the value, and a second prompt to a second large language model resulting in obtaining a second answer, wherein the second large language model comprises the first large language model or another large language model;
communicating the second answer, information about the entity that is separate from the value, and a third prompt to a third large language model resulting in obtaining a third answer, wherein the third large language model comprises the second large language model or another large language model; and
permitting the third answer to be accessed via a device associated with the entity.
16 . The non-transitory computer-readable medium of claim 15 , wherein the value comprises a security value, a sustainability value, a resource efficiency value, a return on investment value, a cost value, or a capacity value.
17 . The non-transitory computer-readable medium of claim 15 , wherein the second large language model is the first large language model, and wherein the third large language model is the first large language model.
18 . The non-transitory computer-readable medium of claim 15 , wherein the first large language model, the second large language model, and the third large language model are different large language models.
19 . The non-transitory computer-readable medium of claim 15 , wherein the second retrieval-augmented generation system is the first retrieval-augmented generation system.
20 . The non-transitory computer-readable medium of claim 15 , wherein the first retrieval-augmented generation system and the second retrieval-augmented generation system are different retrieval-augmented generation systems.