# Based on RAG Model
## Combination of AI Analysis Model and Real-time Data (Based on RAG Model)
Existing AI, especially LLM (Large Language Model), has the following problems:
* Generation of false information (Hallucination)
* Use of outdated information
* Analysis based on data with unclear sources
To solve this, Agendabook introduces a RAG (Retrieval-Augmented Generation) structure. This method works as follows:
* Agendabook collects panel survey response data in real-time
* Records this data on the blockchain to create a "verifiable trusted dataset"
* AI uses this blockchain-based real-time data as the basis (source) for analysis
* As a result, fact-based reliable insights can be derived
This structure preemptively verifies and guarantees data reliability based on blockchain in AI's analysis environment, simultaneously improving reliability, precision, and currency. Additionally, since all data sources are disclosed, it can secure transparency in data analysis, maximizing the advantages of blockchain.
