Projects

Master Thesis - “Exploring the Impact of Extractive Summarization on GPT-based Abstractive Summarization”
GenAI

Study of Transformers-based technologies and creation of a Pipeline with MemSum and Key-BERT for extractive summarization of long documents, evaluating the benefit when conditioning the abstractive summarizer GPT by constraining the abstracts to be generated from sentences that have been extracted either from Memsum or Keybert or possibly both (ensemble-like)

View on GitHub
Smart DOC: a simple RAG Chatbot
GenAI

Developed a Retrieval-Augmented Generation (RAG) chatbot using GPT-4 to dynamically answer queries from uploaded PDF documents. Built with Streamlit (UI), FastAPI/Uvicorn (backend), Pinecone (document indexing), and LangChain for context management.

View on GitHub
Data Mining on flight dataset and ticket price prediction
Data Mining

Dataset analysis, visualization, and preprocessing for subsequent regression and comparison of different models. Random Forest for feature analysis. Stacking ensemble learning method for performance boost

View on GitHub
Neural Network for Multi-class and Multi-label Classification of Movie Trailers
Computer Vision

Training VGG19 and Resnet152 Neural Networks for Genre Recognition from Trailer Frames, Employing Transfer Learning, Fine-tuning, and Data Augmentation.

View on GitHub
OLAP Analysis and Business Intelligence ETL tool usage
BI & Data Warehouse

Definition of a data warehouse system concerning sports betting data, with OLAP analysis conducted using the Pentaho suite and Saiku plugin.

View on GitHub
”Talk2Tag” for vocal annotation of captured photos
Android app development

Developed in Flutter with Dart language and integration with Cloud Storage for Firebase and Cloud Firestore. When the user takes a photo, they can vocally record one or more words to associate with the photo and store them by associating them with it. Later, the app allows searching for photos by one or more words that the user can pronounce.

View on GitHub
AI for Dipole Game
Artificial intelligence and Knowledge Representation

Development of an AI capable of playing Dipole using the Iterative Deepening algorithm, MinMax method, and Alpha–beta pruning technique.

View on GitHub

Education

Master's Degree in Computer Engineering (Artificial Intelligence and Machine Learning)
University of Calabria, Rende
April 2024

•Final Thesis on: Exploring the impact of Extraxtive Summarization on GPT-based Abstractive Summarization

• Relevant Courses: Theoretical Computer Science and Formal Languages, Business Intelligence, Artificial Intelligence and Knowledge Representation, Programming Environments, Architectures and Programming of Processing Systems, Optimization, Machine and Deep Learning, Image and Video Analysis, Distributed Systems and Cloud Computing, Data Mining, Social Network and Media Analysis

Bachelor's Degree in Computer Engineering
University of Calabria, Rende
July 2019

Final Thesis on: Decentralized Identifiers (DIDs) for Self-Sovereign Identity applications

• Relevant Courses: Foundations of Computer Science, Object-Oriented Programming, Algorithms and Data Structures, Computer Architecture, Operating Systems, Databases, Software Engineering, Telecommunication Networks, Software Platforms for Web Applications, Digital Electronics, Electrical Engineering, Fundamentals of Automation, Probabilistic Methods, Operations Research, Mathematical Analysis 1, Mathematical Analysis 2, Physics, Chemistry

Skills

  • Python, Java, C, SQL
  • PyTorch, TensorFlow, Keras, NumPy, scikit‑learn, Pandas, Matplotlib, LangChain, HuggingFace, OpenCV, YOLO, FastAPI, Pentaho
  • Italian (Native), English (Advanced)