AI Scientist Consultant

NEC Laboratories Europe 11/2024 - 02/2025

AI Scientist at NEC Laboratories Europe, within the Biological AI Group.
We developed AI models to predict and generate new virus variants by leveraging their biological composition, and then expanded our approach to analyze and visualize the evolutionary trajectories of virus mutations.
By integrating biological constraints and selective pressures, we refined the evolution process, ensuring that generated variants aligned with plausible evolutionary pathways.
Additionally, we developed novel metrics to assess and refine mutation patterns, further improving predictive capabilities.
TOPICS
- Generative AI virus evolutionary trajectories
- Harmonizing virus evolution through parallel machine learning models processing

Research Intern

NEC Laboratories Europe 04/2024 - 09/2024

Research Intern at NEC Laboratories Europe, within the Biological AI Group. The project aimed to predict and generate new variants of Covid-19 by leveraging the natural evolution of the virus and applying advanced Artificial Intelligence technologies.
We developed a Machine Learning model that incorporated the biological composition of the virus, implementing it into a system capable of recognizing both the virus and its family lineage. Statistical tools were also integrated to enhance the accuracy and predictive power of the model.
TOPICS
- Generative AI models for virus sequences generation
- Reinforcement learning development models for improved virus outbreak prediction

Visiting Researcher Student

The University of Edinburgh 01/2023 - 07/2023

Visiting researcher at the University of Edinburgh's Informatics Forum under the guidance of Dr. Ajitha Rajan.
I contributed to the Horizon 2020 European Project – KATY (Knowledge At the Tip of Your Fingers), focusing on developing AI models for ccRCC (clear cell renal cell carcinoma) prediction. My work involved designing and implementing machine learning training corpora using datasets derived from real human genome sequences.
I developed specialized algorithms to generate these corpora, which included artificial, non-biological proteins and peptides used for machine learning model training. The final distribution of the generated synthetic proteins and peptides closely matched that of the original biological data, ensuring high fidelity to real-world conditions. The generated corpora varied in size, ranging from a few megabytes to several gigabytes.
Additionally, by comparing our results with various benchmark datasets used for training multiple machine learning models, we observed that each generated version exhibited increasingly precise characteristics in both training and test corpora. This refinement improved the overall robustness and accuracy of the models.
I also explored explainability (XAI) models and compositional semantics to enhance model interpretability, ensuring greater transparency in AI-driven predictions.
TOPICS
- Research on Explainable AI and Compositional Semantic Algorithms for ccRCC (clear cell Renal Cell Carcinoma) genome sequences - Corpora generation based on real human genome data

Research fellow

The University of Rome Tor Vergata 01/2021 - 11/2021

First-Class Research Fellow in the project "Automatic Image Captioning for Visual Media Content" at the Department of Enterprise Engineering "Mario Lucertini", University of Rome Tor Vergata.
My work was part of the SFIdA NoW project ("Design and Model Systems for the Training and Information of Operators and Users in Non-Wild Environments"), a collaboration between the University of Rome Tor Vergata and the University of Cassino and Southern Lazio.
The project was funded by INAIL (Istituto Nazionale Assicurazione Infortuni sul Lavoro), Italy’s national institute for workplace accident insurance.
The project aimed to develop an AI model capable of analyzing outdoor sports images collected from social media. Given an image as input, the model assessed safety levels by detecting the presence of personal protective equipment (PPE) and, where applicable, identifying the type of equipment used.
TOPICS
- Image Captioning for Visual Media Contents Research
- Image Classification algorithms

Postgraduate Researcher

The University of Rome Tor Vergata 11/2020 - 01/2020

Postgraduate Researcher in “Interpretability of Neural Network Systems” at the Department of Enterprise Engineering "Mario Lucertini", University of Rome Tor Vergata, in collaboration with AlmaWave, an Italian company specialized in Artificial Intelligence and Natural Language Processing.
The goal of the project was to develop a Conversational Designer (Chatbot) trained on both known datasets and real-world conversations. The objective was to extract valuable information from these conversations and use it as additional knowledge during the training phase of the model, enhancing its performance and improving its ability to interact in a more human-like and accurate manner.
TOPICS
- Interpretability of Neural Network Systems
- Injecting External Knowledge in Conversational AI models

PhD in Artificial Intelligence

Campus Bio-Medico University 11/2021 - 10/2024

PhD Topic: AI-powered prediction of tumor neoantigen burden for personalized cancer vaccines

Master's degree in Computer Science

University of Rome Tor Vergata 10/2018 - 10/2020

Thesis: Hate Speech Models and Where They Fail: Racial and Homophobic Bias on Social Networks

Bachelor's degree in Computer Science

University of Rome Tor Vergata 10/2014 - 05/2018

Thesis: Averaging algorithms for Community Detection: simulation on aleatory and real graphs

Exchange student

King's college London 08/2012 - 09/2012