Granted Patent B2

US 12,602,418 · App. 19/309,601 · Granted Apr 14, 2026

Inventors: Ganesh Prasad Bhat (New Jersey, NJ); James Myers (New York, NY); Zheyu Wang (Shanghai, CN); Haolin Jin (Shanghai, CN); Sourabh Deb (Tampa, FL); Jason Ryan Engelbrecht (London, GB); Payal Jain (London, GB); Tariq Husayn Maonah (London, GB); Mariusz Saternus (Cracow, PL); Daniel Lewandowski (Cracow, PL); Biraj Krushna Rath (London, GB); Stuart Murray (London, GB); Philip Davies (London, GB); Julisia Jackson (Irving, TX); Chamindra Desilva (London, GB); Shardul Malviya (London, GB); Wayne Liao (London, GB); Deepak Jain (London, GB); Samantha Cory (London, GB); Vishal Mysore (Mississauga, CA); Ramkumar Ayyadurai (Jersey City, NJ)

G06F16/338G06F16/383

Abstract

Claims (89)

1 . One or more non-transitory, computer-readable storage medium comprising instructions recorded thereon, wherein the instructions, when executed by at least one data processor of a system, cause the system to:

receive, from a user, a query comprising a request relating to a particular domain;

determine, based on the query, a set of sub-queries using a decomposition model trained to determine sub-queries based on one or more of semantic boundaries, syntactics, tasks, relationships, and rules relating to particular domains;

input the set of sub-queries into a routing model to determine a set of specialized models for the set of sub-queries, wherein the routing model is trained to assign sub-queries for input into a set of specialized models according to one or more routing strategies that balance or prioritize a plurality of factors;

for each particular sub-query in the set of sub-queries:

route the particular sub-query to a respective specialized model in the set of specialized models;

input the particular sub-query into the respective specialized model to generate an output; and

assign, to each respective output, a confidence score based on a reliability of the respective specialized model, a complexity of the particular sub-query, and a relevance of the respective output;

detect a conflict among a set of outputs generated for the set of sub-queries, wherein the conflict is detected using a conflict detection model configured to identify logical, factual, or semantic discrepancies among outputs, and wherein the conflict comprises a discrepancy between two outputs of the set of outputs;

cause the routing model to update, based on the conflict, to minimize future conflicts among sets of outputs generated by the routing model;

generate an aggregated output by combining the set of outputs according to (i) a weighted aggregation algorithm that prioritizes outputs with higher confidence scores and (ii) a plurality of conflict resolution rules, wherein the aggregated output resolves the conflict between the two outputs; and

cause display of the aggregated output in response to the query.

2 . The one or more non-transitory, computer-readable storage medium of claim 1 , wherein the instructions further cause the system to:

extract, from the query, context relating to both the user and a query session involving the user, the context including historical interaction data and relevant system parameters,

wherein the plurality of conflict resolution rules are based on the context relating to both the user and the query session involving the user.

3 . The one or more non-transitory, computer-readable storage medium of claim 1 , wherein the one or more routing strategies comprise one or more of a performance-based routing strategy that prioritizes latency requirements and accuracy thresholds, a cost-optimized routing strategy that balances computational costs with quality requirements, a domain expertise routing strategy that prioritizes expertise of the set of specialized models, and a learning-based routing strategy that updates based on historical performance and real-time feedback.

4 . The one or more non-transitory, computer-readable storage medium of claim 1 , wherein the instructions further cause the system, prior to inputting the set of sub-queries into the routing model, to:

utilize a load balancer, the load balancer being configured to distribute the set of sub-queries across multiple instances of the set of specialized models and manage allocation of system resources to the multiple instances of the set of specialized models,

wherein the load balancer is further configured to prevent resource bottlenecks by directing each sub-query of the set of sub-queries to a selected instance of the set of specialized models based at least in part on a real-time system load and an availability of computational resources.

5 . The one or more non-transitory, computer-readable storage medium of claim 1 , wherein the instructions for determining the set of sub-queries further cause the system to:

identify at least one of:

semantic boundaries or conceptual units within the request;

syntactics within the request, the syntactics comprising clauses, modifiers, or logical operators;

tasks indicated by the request;

entities and relationships between the entities from the request; and

rules specific to components of the request; and

determine the set of sub-queries based on the at least one of the semantic boundaries, the syntactics, the tasks, the entities and the relationships, and the rules.

6 . A method comprising:

receiving, from a user, a query comprising a request relating to a particular domain;

determining, based on the query, a set of sub-queries using a decomposition model trained to determine sub-queries;

inputting the set of sub-queries into a routing model to determine a set of specialized models for the set of sub-queries, wherein the routing model is trained to assign sub-queries for input into a set of specialized models according to one or more routing strategies;

for each particular sub-query in the set of sub-queries:

routing the particular sub-query to a respective specialized model in the set of specialized models;

inputting the particular sub-query into the respective specialized model to generate an output;

assigning, to each respective output, a confidence score; and

classifying one or more intents associated with the query by using a trained intent classification model to detect both a primary intent and at least one secondary intent within the query, wherein the trained intent classification model assigns a confidence score to each detected intent, and

generating an aggregated output by combining the set of outputs according to (i) a weighted aggregation algorithm that prioritizes outputs with higher confidence scores and (ii) a plurality of conflict resolution rules, wherein the aggregated output is based at least in part on the one or more intents; and

causing display of the aggregated output in response to the query.

7 . The method of claim 6 , further comprising:

extracting, from the query, context relating to both the user and a query session involving the user, the context including historical interaction data and relevant system parameters,

wherein the plurality of conflict resolution rules are based on the context relating to both the user and the query session involving the user.

8 . The method of claim 6 , wherein the one or more routing strategies comprise one or more of a performance-based routing strategy that prioritizes latency requirements and accuracy thresholds, a cost-optimized routing strategy that balances computational costs with quality requirements, a domain expertise routing strategy that prioritizes expertise of the set of specialized models, and a learning-based routing strategy that updates based on historical performance and real-time feedback.

9 . The method of claim 6 , further comprising, prior to inputting the set of sub-queries into the routing model:

utilizing a load balancer, the load balancer being configured to distribute the set of sub-queries across multiple instances of the set of specialized models and manage allocation of system resources to the multiple instances of the set of specialized models,

wherein the load balancer is further configured to prevent resource bottlenecks by directing each sub-query of the set of sub-queries to a selected instance of the set of specialized models based at least in part on a real-time system load and an availability of computational resources.

10 . The method of claim 6 , wherein determining the set of sub-queries further comprises:

identifying at least one of:

semantic boundaries or conceptual units within the request;

syntactics within the request, the syntactics comprising clauses, modifiers, or logical operators;

tasks indicated by the request;

entities and relationships between the entities from the request; and

rules specific to components of the request; and

determining the set of sub-queries based on the at least one of the semantic boundaries, the syntactics, the tasks, the entities and the relationships, and the rules.

11 . The method of claim 6 , further comprising:

detecting a conflict among a set of outputs generated for the set of sub-queries, wherein the conflict is detected using a conflict detection model configured to identify logical, factual, or semantic discrepancies among outputs, and wherein the conflict comprises a discrepancy between two outputs of the set of outputs, wherein the aggregated output resolves the conflict between the two outputs; and

causing the routing model to update, based on the conflict, to minimize future conflicts among sets of outputs generated by the routing model.

12 . A system comprising:

a storage device; and

one or more processors communicatively coupled to the storage device storing instructions thereon, that cause the one or more processors to:

receive, from a user, a query comprising a request relating to a particular domain;

determine, based on the query, a set of sub-queries using a decomposition model trained to determine sub-queries;

utilize a load balancer, the load balancer being configured to distribute the set of sub-queries across multiple instances of the set of specialized models and manage allocation of system resources to the multiple instances of the set of specialized models,

wherein the load balancer is further configured to prevent resource bottlenecks by directing each sub-query of the set of sub-queries to a selected instance of the set of specialized models based at least in part on a real-time system load and an availability of computational resources;

input the set of sub-queries into a routing model to determine a set of specialized models for the set of sub-queries, wherein the routing model is trained to assign sub-queries for input into a set of specialized models according to one or more routing strategies;

for each particular sub-query in the set of sub-queries:

route the particular sub-query to a respective specialized model in the set of specialized models;

input the particular sub-query into the respective specialized model to generate an output; and

assign, to each respective output, a confidence score;

generate an aggregated output by combining the set of outputs according to (i) a weighted aggregation algorithm that prioritizes outputs with higher confidence scores and (ii) a plurality of conflict resolution rules; and

cause display of the aggregated output in response to the query.

13 . The system of claim 12 , wherein the instructions further cause the one or more processors to:

extract, from the query, context relating to both the user and a query session involving the user, the context including historical interaction data and relevant system parameters,

wherein the plurality of conflict resolution rules are based on the context relating to both the user and the query session involving the user.

14 . The system of claim 12 , wherein the one or more routing strategies comprise one or more of a performance-based routing strategy that prioritizes latency requirements and accuracy thresholds, a cost-optimized routing strategy that balances computational costs with quality requirements, a domain expertise routing strategy that prioritizes expertise of the set of specialized models, and a learning-based routing strategy that updates based on historical performance and real-time feedback.

15 . The system of claim 12 , wherein the instructions for determining the set of sub-queries further cause the one or more processors to:

identify at least one of:

semantic boundaries or conceptual units within the request;

syntactics within the request, the syntactics comprising clauses, modifiers, or logical operators;

tasks indicated by the request;

entities and relationships between the entities from the request; and

rules specific to components of the request; and

determine the set of sub-queries based on the at least one of the semantic boundaries, the syntactics, the tasks, the entities and the relationships, and the rules.

16 . The system of claim 12 , wherein the instructions further cause the one or more processors to:

classify one or more intents associated with the query by using a trained intent classification model to detect both a primary intent and at least one secondary intent within the query,

wherein the trained intent classification model assigns a confidence score to each detected intent, and

wherein the aggregated output is based at least in part on the one or more intents.

17 . The system of claim 12 , wherein the instructions further cause the one or more processors to:

detect a conflict among a set of outputs generated for the set of sub-queries, wherein the conflict is detected using a conflict detection model configured to identify logical, factual, or semantic discrepancies among outputs, and wherein the conflict comprises a discrepancy between two outputs of the set of outputs, wherein the aggregated output resolves the conflict between the two outputs; and

cause the routing model to update, based on the conflict, to minimize future conflicts among sets of outputs generated by the routing model.

Assignments (1)

Continuity (14)

References Cited (326)