Systems and methods for providing advanced personalization in query systems

1. A computing system comprising:

a processor; and

a non-transitory computer-readable storage device storing computer-executable instructions, the instructions operable to cause the processor to perform operations comprising:

receiving a user query from a user device;

collecting historical data for a user associated with the user query;

formatting the historical data and the user query as an input prompt;

feeding the input prompt to a large language model (LLM);

generating one or more tuning parameters via the LLM by analyzing the input prompt;

updating a user profile of the user based on the one or more tuning parameters;

identifying one or more matching user queries based on the received user query;

generating, via a prediction engine, one or more personalized recommendations in response to the user query;

updating the input prompt with the one or more personalized recommendations;

feeding the updated input prompt to the LLM;

generating, via the LLM, one or more final responses to the user query by analyzing the updated input prompt; and

causing the one or more final responses to be displayed on the user device.

2. The computing system of claim 1 , wherein collecting the historical data for the user comprises collecting at least one of one or more past queries from the user, a plurality of interaction data for the user, or feedback data for the user.

3. The computing system of claim 1 , wherein collecting the historical data for the user comprises collecting the user profile of the user, the user profile comprising a knowledge graph.

4. The computing system of claim 1 , wherein generating the one or more tuning parameters via the LLM comprises generating one or more of an inferred preference, an inferred expertise area, a query pattern, or a resource use pattern.

5. The computing system of claim 1 , wherein identifying the one or more matching user queries based on the received user query comprises performing a hybrid semantic search on a database.

6. The computing system of claim 1 , wherein generating the one or more personalized recommendations in response to the user query comprises:

identifying one or more similar users to the user;

constructing a user-item interaction matrix for the user;

decomposing the user-item interaction matrix; and

calculating a predicted interaction of the user for an item.

7. The computing system of claim 1 , wherein causing the one or more final responses to be displayed on the user device comprises at least one of:

displaying a ranked list of resources;

displaying a confidence score associated with each final response; or

displaying an explanation why at least one final response was suggested.

8. The computing system of claim 1 , wherein the operations comprise:

receiving interaction data for the one or more final responses;

causing a feedback option to be displayed on the user device;

receiving feedback data from the user; and

retraining the prediction engine based on the interaction data and the feedback data.

9. The computing system of claim 8 , wherein receiving the interaction data comprises receiving an indication of the user following a link, opening a document, or utilizing a code snippet from the one or more final responses.

10. The computing system of claim 8 , wherein receiving the feedback data comprises receiving a selection of an upvote, a selection of a downvote, or textual feedback.

11. A computer-implemented method, performed by at least one processor, comprising:

receiving a user query from a user device;

collecting historical data for a user associated with the user query;

formatting the historical data and the user query as an input prompt;

feeding the input prompt to a large language model (LLM);

generating one or more tuning parameters via the LLM by analyzing the input prompt;

updating a user profile of the user based on the one or more tuning parameters;

identifying one or more matching user queries based on the received user query;

generating, via a prediction engine, one or more personalized recommendations in response to the user query;

updating the input prompt with the one or more personalized recommendations;

feeding the updated input prompt to the LLM;

generating, via the LLM, one or more final responses to the user query by analyzing the updated input prompt; and

causing the one or more final responses to be displayed on the user device.

12. The computer-implemented method of claim 11 , wherein collecting the historical data for the user comprises collecting at least one of one or more past queries from the user, a plurality of interaction data for the user, or feedback data for the user.

13. The computer-implemented method of claim 11 , wherein collecting the historical data for the user comprises collecting the user profile of the user, the user profile comprising a knowledge graph.

14. The computer-implemented method of claim 11 , wherein generating the one or more tuning parameters via the LLM comprises generating one or more of an inferred preference, an inferred expertise area, a query pattern, or a resource use pattern.

15. The computer-implemented method of claim 11 , wherein identifying the one or more matching user queries based on the received user query comprises performing a hybrid semantic search on a database.

16. The computer-implemented method of claim 11 , wherein generating the one or more personalized recommendations in response to the user query comprises:

identifying one or more similar users to the user;

constructing a user-item interaction matrix for the user;

decomposing the user-item interaction matrix; and

calculating a predicted interaction of the user for an item.

17. The computer-implemented method of claim 11 , wherein causing the one or more final responses to be displayed on the user device comprises at least one of:

displaying a ranked list of resources;

displaying a confidence score associated with each final response; or

displaying an explanation why at least one final response was suggested.

18. The computer-implemented method of claim 11 comprising:

receiving interaction data for the one or more final responses;

causing a feedback option to be displayed on the user device;

receiving feedback data from the user; and

retraining the prediction engine based on the interaction data and the feedback data.

19. The computer-implemented method of claim 18 , wherein receiving the interaction data comprises receiving an indication of the user following a link, opening a document, or utilizing a code snippet from the one or more final responses.

20. The computer-implemented method of claim 18 , wherein receiving the feedback data comprises receiving a selection of an upvote, a selection of a downvote, or textual feedback.