LLM Chatbot Function Selection for Lower-Bandwidth App Building

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing computer applications rely heavily on manual selection and sequencing of commands and functions by programmers, which is inefficient and can lead to increased bandwidth usage and processing overhead when interacting with large language model chatbots.

Innovation Solution

A system that processes user requests to identify context and embeddings, filters available functions, and interacts with a large language model chatbot to determine the necessary functions and their execution order, reducing data transmission and distributing processing tasks between the system and the chatbot.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual selection and sequencing of commands and functions is used by programmers, then application functionality is achieved, but development efficiency is reduced and processing overhead increases

Engineering Contradiction:

Improveapplication development efficiencyVSAvoidprogrammer time for command selection

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system enables self-service by allowing the LLM chatbot to automatically select and sequence commands and functions based on user intent, eliminating the need for manual programmer intervention in command selection while maintaining application functionality

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical process of manual command selection with an intelligent system using LLMs that can understand natural language and automatically determine the appropriate sequence of commands and functions to execute

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If comprehensive function descriptions are transmitted to the large language model chatbot, then the chatbot can accurately determine necessary functions, but bandwidth usage increases

Engineering Contradiction:

Improvefunction identification accuracyVSAvoidbandwidth consumption

Core Design Contradiction:

Measurement precisionVSLoss of energy

Solution Approach 1:

The system extracts only the essential function signatures and descriptions needed for the LLM to make accurate decisions, removing redundant information while maintaining sufficient context for precise function identification

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transmits a curated subset of function information that is sufficient for accurate function determination without providing complete exhaustive descriptions, achieving the right balance between accuracy and bandwidth efficiency

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS20250216955A1System and method for building computer applications using large language model chatbots

Publication Date: 2025.07.03 CVENT INC

AI summary

Systems, methods, and computer-readable storage media for building computer applications, and more specifically to building computer applications where the commands and functions of the computer applications are determined using a Large Language Model (LLM) chatbot. A system can receive a question, and determine the context of the question. The system can then transmit the question with the context to a large language model chatbot. The system can then receive, at the computer system from the large language model chatbot based on the question and the context, at least one function. The system can execute the function(s), and send the result back to the chatbot. This process may repeat, with the end result being an answer to the original question, where the answer is generated by the chatbot.