Education
M.S. Artificial Intelligence and Data Engineering
University of Pisa, 2020-2022
In the field of Artificial Intelligence, I've explored topics like Data Mining, Machine Learning, Computer Vision, Natural Language Processing, Optimization Theory, and Process Mining. On the other hand, in Data Engineering, I've gained hands-on experience managing large databases, learned about Distributed Systems and Cloud Computing, and developed skills in areas like IoT devices and Business/Project Management. Throughout this journey, I e become skilled in using various tools, including MySQL, LevelDB, MongoDB, Neo4j, Git, Docker, Kubernetes, Tensorflow, and Pytorch.
- Final grade: 110/110 summa cum laude (4.0 GPA)
- Thesis published in a Tier A conference
- Exams and Marks link
B.S. Computer Engineering
University of Pisa, 2017-2020
I gained a strong foundation in engineering, with a particular focus on computer engineering. This included in-depth knowledge of mathematics, physics, algorithms, databases, computer architecture, computer networks, operating systems, and programming languages such as C, C++, Java, Python, Matlab, SQL, JavaScript, and PHP.
- Final grade: 110/110 (4.0 GPA)
- Exams and Marks link
Scientific High School Diploma
ISS P. Aldi - Liceo Scientifico G. Marconi, 2012-2017
- Final grade: 100/100 (4.0 GPA)
Experience
Translated - AI Engineer & Researcher
October 2023 - PRESENT
- Language Expansion: I expanded the number of supported languages by our MT from 56 to 201, making it the first commercial translation engine in the world with such extensive language support Learn more here
- Trust Attention: I had the intuition for a novel technique that prioritizes the most valuable training data. This approach played a pivotal role in significantly enhancing the quality of machine translation, resulting in improvements unseen in the past five years. Learn more here
- LLM-based translation model: I was chosen to be part of a new team specifically dedicated to build the next generation of machine translation with a different paradigm based on Large Language Model
- Polyglot: I expanded the language support of the Language Identification model from 56 to 201 languages, all while maintaining the same level of accuracy and inference time.
Published Research
The Emotions of the Crowd: Learning Image Sentiment from Tweets via Cross-modal Distillation
ECAI 2023 - 26th European Conference on Artificial Intelligence - Tier A conference
April 2023
In our paper, we present an innovative multi-modal distillation framework designed to address the challenge of sentiment analysis. This framework has demonstrated remarkable performance, surpassing the current state of the art in five distinct benchmark tests. Notably, our approach involves transferring knowledge from a high-accuracy textual model to a visual model, enabling the accurate assignment of sentiments to images. As part of this research, we've also compiled a substantial dataset, comprising approximately 1.5 million images gathered from Twitter over a span of three months. We've taken a collaborative approach by releasing both the model weights and the dataset as open-source resources, thereby fostering future research opportunities in this domain. Official paper website.
Honors & Awards
2° place at Loop Q Prize AI competition
June 2022
LoopQPrize is an European and African competition, which focuses on the fields of Cognitive Computing and Machine Learning. The 2022 edition involved the training of at least 5 models able to recognize the prevailing emotion present in the speech recordings among seven possible categories (angry, happy, disgust, fear, sadness, neutral and surprise). The submitted solution was evaluated considering the innovations in the technique used, the correctness and completeness of the solution, the prediction performance and the scalability potential. github.com link.
Skills
Note: I think these sections are silly, but everyone seems to have one. Here is a mostly honest overview of my skills.
Selected Courses
696AA:
Optimization Methods and Game Theory
•
793AA:
Cloud Computing
•
875II:
Business and Project Management
•
875II:
Computational Intelligence and Deep Learning
•
878II:
Data Mining and Machine Learning
•
879II:
Distributed Systems and Middleware Technologies
•
882II:
Internet of Things
•
883II:
Large Scale and Multi-Structured Databases
•
886II:
Multimedia Information Retrieval and Computer Vision
•
888II:
Process Mining and Intelligence
•
893II:
Symbolic and Evolutionary Artificial Intelligence
