Chatbot with LLM and Vectorized Database for Card Benefit Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Consumers often fail to understand the benefits available to them through various payment cards in their wallet, nor which merchants provide specific benefits unique to one card over another.
Innovation Solution
An apparatus and method utilizing a processor and memory to store credit card documentation, converse with users via a chatbot, identify benefits, execute a large language model (LLM) to generate confirmation letters, and transmit these via electronic messages to user devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a chatbot with LLM and vectorized database is implemented to provide personalized card benefit information, then user understanding and utilization of benefits is improved, but system complexity and computational resources increase
Solution Approach 1:
The patent introduces a chatbot as an intermediary between the user and the complex payment card benefit system. The chatbot translates user queries into structured questions, interacts with the vectorized database to retrieve relevant information, and presents results in natural language. This intermediary layer simplifies the user interface while managing the underlying system complexity through standardized interaction protocols and pre-processed knowledge representations.
Solution Approach 2:
The system performs preliminary actions by pre-processing and vectorizing card documentation before runtime queries. The vectorized database stores structured representations of benefit information, terms, and conditions in advance. When users query the chatbot, the system retrieves pre-processed information through similarity matching rather than parsing raw documents in real-time, significantly reducing computational complexity during user interactions.
2Ease of operation
If natural language processing and LLM execution are used to generate personalized responses, then user experience and information accuracy are improved, but processing time and computational power increase
Solution Approach 1:
The chatbot implements partial action by selectively processing only the most relevant portions of user queries and matching them against corresponding sections of the vectorized database. Rather than analyzing entire documents or generating completely new responses for every query, the system retrieves pre-processed information segments and combines them to answer specific questions, reducing processing time while maintaining personalized and accurate responses.
3Measurement precision
If comprehensive card documentation is stored and analyzed, then benefit identification accuracy is improved, but data storage requirements and processing load increase
Solution Approach 1:
The system extracts essential information from comprehensive card documentation and stores only the critical elements in the vectorized database. Rather than storing and processing entire card terms and conditions, the system extracts key benefit definitions, eligibility criteria, and reward structures into structured representations. This extraction process reduces data storage requirements while maintaining the precision needed for accurate benefit identification through targeted similarity matching.
Data Source
AI summary
An example operation may include one or more of receiving an input from a user during a conversation that includes a plurality of prompts between the user and a chatbot within a chat window of a software application, converting text content within the received input into a vector, executing a large language model (LLM) on the vector and a database of vectorized responses to identify a vectorized response to output from among the plurality of vectorized responses within the database, converting the vectorized response into a text response, and displaying the text response output by the chatbot within the chat window of the software application.