• Ongoing PhD in Information Technology (2024 - present), Politecnico di Milano, Milan, Italy.
• MSc in Computer Science and Engineering (2020 - 2023), Politecnico di Milano, Milan, Italy.
• BSc in “Ingegneria Informatica” (2017 - 2020), Politecnico di Milano, Milan, Italy.
• Scientific High School Diploma (2012 - 2017), Liceo Scientifico "G. Galilei", Tarquinia, Italy.

Title:Machine learning-based forecast of Helmet-CPAP therapy failure in acute distress respiratory syndrome patients
Supervisor: Prof. M. Masseroli
Abstract: This study uses machine learning to predict the failure of Helmet Continuous Positive Airway Pressure (H-CPAP) therapy in patients with Acute Respiratory Distress Syndrome (ARDS). Data from Vimercate Hospital in Italy is used to create a comprehensive pipeline, including data gathering, preprocessing, and cleaning. The best model achieves 92.7% accuracy. Key clinical features influencing predictions, identified through Shapley additive explanations, include oxygen levels, C-reactive protein, heart rate, respiratory rate, and others. The results suggest that interpretable machine learning can aid early clinical decision-making for ARDS patients undergoing this therapy.

• Artificial Intelligence: Artificial NN and Deep Learning, Uncertainty in AI, Streaming Data Analytics.
• Computer Science: Soft. Eng. 2, DBs 2, Distr. Sys., Adv. Comp. Arch., Formal Langs. & Compilers.
• Computer Security: Computer Security, Digital Forensics and Cybercrime.
• Other: Computer Graphics, Mobile Applications Development.

• A. De Santis*, R. Campi*, M. Bianchi and M. Brambilla, "Visual-TCAV: Concept-based Attribution and Saliency Maps for Post-hoc Explainability in Image Classification", Transactions on Machine Learning Research (TMLR), 2835-8856, 2026.

• R. Campi, A. De Santis, P. Colombo, P. Scarpazza and M. Masseroli, "Machine learning-based forecast of Helmet-CPAP therapy failure in Acute Respiratory Distress Syndrome patients", Computer Methods and Programs in Biomedicine (CMPB), vol. 260, 2025, 108574, ISSN 0169-2607.

• R. Campi, M. Giudici, N. O. Pinciroli Vago, M. Brambilla and P. Fraternali, "Enhancing Human-AI Collaboration through a Conversational Agent for Energy Efficiency", Proceedings of the 2025 AAAI Spring Symposium Series, vol. 5, no. 1, pp. 52-55, May 2025.

• G. Astolfi*, M. Bianchi*, R. Campi*, A. De Santis and M. Brambilla, "A Framework for Evaluating Zero-Shot Image Generation in Concept-based Explainability", accepted at 2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition Findings (CVPRF), Denver, CO, USA, 2026.

• M. Bianchi*, R. Campi*, A. De Santis*, S. Merengo* and M. Brambilla, "Activation-Based Concept Extraction for Explainability in Image Classification", accepted at 2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Denver, CO, USA, 2026.

• R. Campi, S. Borrego, A. De Santis, M. Bianchi, A. Tocchetti and M. Brambilla, "Towards Synthetic Concept Activation Vectors via Generative Models", 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Nashville, TN, USA, 2025, pp 2711-2719.

• R. Campi, N.O. Pinciroli Vago, M. Giudici, P.B. Rodriguez‑Guisado, M. Brambilla and P. Fraternali, "A Graph-Based RAG for Energy Efficiency Question Answering", 2025 International Conference on Web Engineering (ICWE), Lecture Notes in Computer Science, vol 15749, Springer, Cham, 2025.

• A. De Santis*, R. Campi*, M. Bianchi*, A. Tocchetti and M. Brambilla, "Foundational approaches to post-hoc explainability for image classification", Bi-directionality in Human-AI Collaborative Systems, Academic Press, 2025, Pages 23-54, ISBN 9780443405532.

• A. Tocchetti, M. Bianchi*, R. Campi*, A. De Santis* and M. Brambilla, "On the principles and effectiveness of gamification in bidirectional artificial intelligence and explainable AI", Bi-directionality in Human-AI Collaborative Systems, Academic Press, 2025, Pages 135-160, ISBN 9780443405532.

• R. Campi, N. O. Pinciroli Vago, M. Giudici, M. Brambilla and P. Fraternali, "MDER-DR: Multi-Hop Question Answering with Entity-Centric Summaries", arXiv:2603.11223, 2026.

• Explainable AI (XAI) and Mechanistic Interpretability (MI):
  Dissecting internals of deep networks such as LLMs, Vision Transformers (ViTs) and CNNs.

  • Concept-based Explainability:
    Designing algorithms that use human-understandable visual concepts to explain how image models ground their decisions.

  • Sparse and Interpretable Architectures:
    Designing architectures such as Sparse Autoencoders (SAEs), Transcoders, and Concept Bottleneck Models (CBMs).

  • FFT-based Anomaly and Drift Detection:
    Analyzing spectral, sparse signatures of generated text for early warning of model failure (e.g., hallucination).

• Knowledge Graphs (KGs), Semantic Web, and Retrieval Augmented Generation (RAG):
  Semantic Web and KGs, RAG pipelines such as indexing and retrieval, and Multi-Hop Question Answering (MH-QA).

• Biomedical AI:
  Predictive modeling for ARDS patients, clinical dataset curation, and interpretability in healthcare applications.

• 2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026):
  Poster preparation and presentation for "A Framework for Evaluating Zero-Shot Image Generation in Concept-based Explainability",
  Poster presentation for "Activation-Based Concept Extraction for Explainability in Image Classification",
  Denver, CO, USA, 2026.

• 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2025):
  Poster preparation for "Towards Synthetic Concept Activation Vectors via Generative Models",
  Nashville, TN, USA, 2025.

• 2025 International Conference on Web Engineering (ICWE 2025):
  Oral presentation of "A Graph-Based RAG for Energy Efficiency Question Answering",
  Delft, Netherlands, 2025.

• 2025 AAAI Spring Symposium on Bi-directionality in Human-AI Collaborative Systems:
  Oral and poster presentation of "Enhancing Human-AI Collaboration through a Conversational Agent for Energy Efficiency",
  Oral and poster presentation of "Human-AI Collaboration in the Fashion Design Process",
  San Francisco, CA, USA, 2025.

• 2025 International Semantic Web Research Summer School (ISWS 2025):
  Oral and poster presentation on Kg-based RAG for Energy Efficiency,
  Bertinoro, Italy, 2025.

• AI Interpretability Seminar (June 19th, 2026):
  Seminar on "From Concept-based Explainable AI for CNNs to Mechanistic Interpretability for LLMs",
  University of Illinois Chicago, USA.

• Explainable AI Seminar (April 23th, 2026):
  Seminar on "Past, Present, and Future of AI Interpretability",
  Xi'an Jiaotong University, China.

• Data Science Seminar on Explainable AI (April 9th, 2025):
  Seminar on "Foundations of Explainable AI: Principles and Model-Agnostic Approaches to Transparency",
  Politecnico di Milano, Italy.

• ICML 2026 Gold Reviewer Award:
  Recognized among top reviewers and awarded complimentary conference registration.

• Conference Reviewer:
  ICML 2026, CVPR 2026, AISTATS 2026, AAAI 2026, ICWE 2025, AAAI SSS 2025, ICML 2025, CVPR 2025, NeurIPS 2024

• Energenius Project:
  Task contributor in the Energenius project (European Union’s Horizon Europe), focused on developing a conversational agent for energy efficiency.

• Website Maintainer:
  Maintaining the Data Science Lab website, curating content and resources for students and researchers.

• 2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026)
  Denver, CO, USA, 2026.

• 2025 International Conference on Web Engineering (ICWE 2025)
  Delft, Netherlands, 2025.

• 2025 AAAI Spring Symposium on Bi-directionality in Human-AI Collaborative Systems
  San Francisco, CA, USA, 2025.

• 2024 European Conference on Computer Vision (ECCV 2024)
  Milan, Italy, 2024.

• Complex Networks: Theory, Methods, and Applications (2024)
  Politecnico di Milano, Como, Italy, 2024.

• 2025 International Semantic Web Research Summer School (ISWS 2025)
  Bertinoro, Italy, 2025.

• Reply S.r.l. Code Challenge "Prompting the Future" (Nov 30th, 2024):
  Judge for the competition as Data Science Lab Representative,
  Milan, Italy.

• BEST Academy Workshop (BEST Milan & Levels Tech) (Oct 31st, 2024):
  Workshop on "Document Intelligence" as Data Science Lab Representative,
  Milan, Italy.

• Massachusetts Institute of Technology, USA:
  Providing mentorship to a MIT researcher on synthetic concept generation for XAI.

• Vimercate Hospital, Italy:
  Collaboration for the development of a machine learning-based predictive model for ARDS patients treated with Helmet-CPAP.

• SUPSI University, Switzerland:
  Collaboration for the development of a graph-based RAG system for energy efficiency.

• Università di Pisa, Italy:
  Collaboration for the development of analytical tools for assessing the impact of conservation treatments on textiles.

• Polimi Data Scientists, Italy:
  Collaboration with "Polimi Data Scientists" MSc students on various research projects.

• BEST (Board of European Students of Technology) Milan, Italy:
  Collaboration with BEST Milan on workshops and talent competition.

• Web and Data Science Course Projects (2025-2026):
  Supervised Web and Data Science course projects for MSc students at Politecnico di Milano.

• Synthetic Concept Generation for XAI (2024-2025):
  Mentored a visiting student from MIT at Politecnico, funded under the Rocca Project, focusing on synthetic concept generation for XAI.

• Polimi Data Scientists (2024-2025):
  Supervised "Polimi Data Scientists" MSc students on various research projects, including RAG, KGs, and XAI.

• Explainable AI and Mechanistic Interpretability:

  • Giacomo Astolfi (2025):
    "A closer look at text-to-image generation for concept-based explainability".

  • Sara Cavallini:
    ongoing.

  • Alessandro Cogollo:
    ongoing.

  • Daniele Di Santi (2025):
    "Concept-based explanations for image classifiers using textual prompts".

  • Tommaso Giordano (2025):
    "Discovering sparse concept graphs for mechanistic interpretability in vision transformers".

  • Martina Maiorana:
    ongoing.

  • Sara Merengo (2024):
    "Activation-Based Concept Extraction for Post-hoc Explainability of CNN Models".

  • Pierluigi Porri (2025):
    "Addressing scale-invariance and missing context in concept-based explainability for CNNs".

  • Giovanni Visi,
    ongoing.

• Knowledge Management, KGs, and RAG:

  • Matteo Angelini (2025):
    "Benchmarking Commercial and Ad-Hoc RAG Implementations in the Energy Domain".

  • Michele Bersani (2026):
    "A Graph-RAG Architecture for Retrieval of Tax Legislation and Administrative Circulars based on Domain Ontology and Citation graphs".

  • Alfredo Ceci (2025):
    "Property Graph RAG: An LLM-driven Approach".

  • Davide Morelli (2026):
    "MathRAG: An Agentic Question Answering Architecture for Mathematical Problems based on Structured Knowledge Graphs".

  • Martina Riva (2026):
    "LLM-based Knowledge Graph Construction for the Legal Sector: Modeling Semantics, Document Structure, and Legal References".

  • Zhenzhen Shan:
    ongoing.

• Biomedical AI:

  • Alessia Marino (2025):
    "Modeling HCPAP therapy outcome prediction in ARDS patients via large language models".

• Systems and Methods for Big and Unstructured Data (2025-2026):
  5 CFUs, MSc, Politecnico di Milano, Italy.

• Web and Data Science (2025-2026):
  5 CFUs, MSc, Politecnico di Milano, Italy.

• Data Science course (2025-2026):
  BSc, Xi'an Jiaotong University under the XJTU–POLIMI Joint School collaboration, China.

• Prova Finale di Ingegneria del Software (2024-2025):
  3 CFUs, BSc, Politecnico di Milano, Italy.

• Informatica B (2024-2025):
  7 CFUs, BSc, Politecnico di Milano, Italy.

• Private tutoring for various computer science topics (2024-2025):
  BSc and MSc students, Lombardy, Italy.

• ENI S.p.A. (November 2025):
  Held a 4 hours course on Knowledge Management, Knowledge Graphs, and Retrieval Augmented Generation for the employees of ENI, Milan, Italy.

• Keysight Technologies S.r.l. (February 2025):
  Held a 16 hours course on Explainable Artificial Intelligence for the employees of Keysight Technologies, Milan, Italy.

• RCS Media Group S.r.l. (July 2024):
  Held a 4 hours course on Advanced Python Programming for the employees of RCS Media Group, Milan, Italy.

• Research Fellow, Politecnico di Milano (2023 - 2024):
  Conducting research in Explainable AI, Mechanistic Interpretability, Knowledge Graphs, RAG, and Biomedical AI; teaching courses; supervising theses; collaborating on research projects.

• Webmaster & Social Media Manager, Optical Expert S.r.l. (2016 - 2020):
  Managed the company website, social media accounts, and online marketing efforts.

• Ambulance Transport Operator, Croce d'Oro Milano ONLUS (2023-2024):
  Provided emergency transport services for patients in need.

• Background Actor, Lotus Production S.r.l. (July - August 2019):
  Participated as a background actor in the film Gli anni più belli directed by Gabriele Muccino.

• Football Referee, AIA Civitavecchia, FIGC (2016 - 2017):
  Officiated football matches for the U14 PROVINCIAL C11 MALE category.

• Italian: Native proficiency.

• Spanish: Fluent (family background).

• English: Professional working proficiency (TOEIC certified).

• AI Frameworks:
  PyTorch, TensorFlow, Hugging Face Transformers.

• Knowledge Graph and Semantic Web:
  RDFLib, SPARQL, Neo4j.

• RAG and LLM Stack:
  Ollama, LangChain, embedding pipelines, vector & graph search, ChromaDB.

• Data Science:
  Pandas, NumPy, Scikit‑learn, Matplotlib, Seaborn, SciPy.

• Tools:
  Jupyter, Docker, Git, LaTeX, draw.io, VS Code, Linux/Unix environments.