Document management and analysis are fundamental processes in modern companies, often hindered by manual, repetitive, and costly tasks. This thesis describes the development and implementation of a software solution that automates and optimizes the analysis of various types of business documentation. The proposed approach is based on the use of existing, validated documents as reference "templates". Through the use of Large Language Models (LLMs) and embeddings, the system processes new documentation at both semantic and structural levels, classifying and comparing it against the reference models to evaluate its compliance and content consistency.
La gestione e l’analisi della documentazione rappresentano processi fondamentali in un'azienda moderna, spesso rallentati da attività manuali, ripetitive e costose. Questa tesi descrive lo sviluppo e l’implementazione di un software che automatizza e ottimizza l’analisi di diverse tipologie di documentazione aziendale. L'approccio proposto si basa sull’impiego di documenti esistenti e validati, utilizzati come "template" di riferimento. Grazie all'impiego di Large Language Models (LLM) ed embeddings, il programma elabora la nuova documentazione a livello semantico e strutturale, classificandola e confrontandola con i modelli di riferimento per valutarne la conformità e la coerenza dei contenuti.
AI Data Supervisor - Supervisione intelligente della documentazione aziendale
BERTI, DAMIANO
2025/2026
Abstract
Document management and analysis are fundamental processes in modern companies, often hindered by manual, repetitive, and costly tasks. This thesis describes the development and implementation of a software solution that automates and optimizes the analysis of various types of business documentation. The proposed approach is based on the use of existing, validated documents as reference "templates". Through the use of Large Language Models (LLMs) and embeddings, the system processes new documentation at both semantic and structural levels, classifying and comparing it against the reference models to evaluate its compliance and content consistency.| File | Dimensione | Formato | |
|---|---|---|---|
|
Berti_Damiano.pdf
Accesso riservato
Dimensione
2.47 MB
Formato
Adobe PDF
|
2.47 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/116475