LLM Stream Analysis With Event Rate Reduction for IoT Data

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current large language model (LLM)-based agents exhibit query response times in seconds rather than milliseconds or microseconds, making them unsuitable for near real-time analysis of event streams, whether from a single heterogeneous stream or multiple streams.

Innovation Solution

A stream analysis system that processes incoming data streams to reduce the event rate to a level compatible with LLMs by dynamically determining response times and applying event rate reduction strategies, such as sampling, aggregating, and correlating events, to generate an adjusted data stream for analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If LLM-based agents are used for data stream analysis, then analysis accuracy is improved, but response time increases to seconds instead of milliseconds

Engineering Contradiction:

Improveanalysis accuracyVSAvoidresponse time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the data stream processing into multiple stages: event pre-processing, filtering, aggregation, and selective LLM analysis. Only significant events are forwarded to the LLM, reducing the number of queries while maintaining analysis accuracy for critical events.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies partial action by using simplified processing rules for routine events and reserving LLM analysis for exceptional or high-value events. This selective approach reduces overall response time while maintaining accuracy where needed.

Inventive Principle:

Principle #16Partial or excessive action

2Productivity

If event rate reduction strategies are applied, then LLM processing compatibility is improved, but information loss may occur

Engineering Contradiction:

Improveprocessing compatibilityVSAvoidevent information

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The system performs preliminary actions by pre-processing events before LLM analysis, including filtering out low-value events, aggregating related events, and extracting key features. This prepares the data stream for efficient LLM processing while preserving critical information.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary layer between the data stream and LLM that acts as a mediator. This layer includes event processors, filters, and aggregators that transform raw events into condensed representations, reducing information loss while enabling LLM compatibility.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20260025429A1Analyzing data streams using large language models

Publication Date: 2026.01.22 VANTIQ INC

AI summary

A system and method comprising receiving a data stream including a plurality of events from sensors within an Internet of Things (IoT) computing environment, applying a rate reduction method to generate an adjusted data stream by adjusting a rate of the data stream, the rate reduction method being based on an estimated response time for a stream analysis model, and transmitting the adjusted data stream to the stream analysis model. The system further updates the estimated response time for the stream analysis model based on receiving, from the stream analysis model, an output based on an event from the adjusted data stream. The system stores the event and the output, and transmits the event and the output to an automatic control system.