Tesi etd-03252026-153225 |
Link copiato negli appunti
Tipo di tesi
Tesi di laurea magistrale
Autore
ANDRIANI, PAOLO
URN
etd-03252026-153225
Titolo
Design and implementation of a streamset ETL pipeline supported by Generative AI
Dipartimento
INFORMATICA
Corso di studi
DATA SCIENCE AND BUSINESS INFORMATICS
Relatori
relatore Prof. Mencagli, Gabriele
tutor Bardone, Andrea
tutor Bardone, Andrea
Parole chiave
- AI
- ETL Migration
- LLMs
Data inizio appello
10/04/2026
Consultabilità
Non consultabile
Data di rilascio
10/04/2066
Riassunto (Inglese)
This thesis presents the design and implementation of an AI-based multi-agent system for the automated analysis, documentation, and migration of legacy data processing pipelines used at Reale Mutua.
The proposed system includes an agent that analyzes IBM DataStage .dsx files to generate structured JSON documentation and a second agent that converts this information into equivalent Python code for migration to modern environments.
Additional agents support legacy modernization by generating Python code from technical documentation and extracting structured information from legacy Java insurance product logic.
The results show that AI-driven agents can significantly reduce manual effort, improve maintainability, and facilitate the transition to modern data engineering architectures.
The proposed system includes an agent that analyzes IBM DataStage .dsx files to generate structured JSON documentation and a second agent that converts this information into equivalent Python code for migration to modern environments.
Additional agents support legacy modernization by generating Python code from technical documentation and extracting structured information from legacy Java insurance product logic.
The results show that AI-driven agents can significantly reduce manual effort, improve maintainability, and facilitate the transition to modern data engineering architectures.
Riassunto (Italiano)
File
| Nome file | Dimensione |
|---|---|
La tesi non è consultabile. |
|