About Me
I am an Italian-American-New-York-Mainer-Inter-Milan-Supporter. My family comes from Friuli and Istria, where there are many endangered languages and lots of efforts to preserve and revitalize them. I try to help by making tools that speed up language documentation.
Research
I am a master's student in Data Science and Artificial Intelligence at the University of Trieste. I graduated from New York University with a B.A. in Mathematics and Computer Science, where I was a Shelby Davis Scholar. My research focuses on developing computational tools for low-resource and under-resourced languages, with the goal of making language technologies more inclusive and practically useful for linguistic research and documentation.
My work centers on automatic speech recognition (ASR) and related speech technologies, including forced alignment, speaker diarization, segmentation, and part-of-speech tagging in low-resource settings. I am particularly interested in linguistically informed modeling choices—such as phonological and featural representations—and how these choices affect model performance, generalization, and usability in real-world documentation pipelines. I belive that computational approaches to language preservation must be community-oriented.
Publications & Presentations
-
Linguistically Informed Tokenization Improves ASR for Underresourced Languages
Field Matters, 2026 — First Author -
NLP Tools for Constructing a Spoken Corpus of Endangered Language Varieties: A Case Study of the ELIC Corpus
Presented at AACL, 2026 — First Author -
Comparing Phonological Feature Sets for Low-Resource ASR
Proceedings of the 2025 Annual Meeting on Phonology — Second Author -
Forced Alignment of Istro-Venetian Speech via Neural versus Non-Neural Models
The Journal of the Acoustical Society of America, 159, A161, 2026 — Second Author -
Creating a Multilingual Spoken Corpus for Language Documentation and Research
Presented at ICLDC — Co-Author
Research Experience
-
Computational Intern – Low-Resource Language Documentation
Endangered Languages in Contact, Istria & Kvarner / New York City (06/2024 – Present)
I lead projects fine-tuning custom Wav2Vec2-BERT models for automatic speech recognition and forced alignment of low-resource languages. I also develop machine learning approaches for speaker diarization, segmentation, and part-of-speech tagging to support corpus creation and linguistic analysis. -
Computational Intern – Addiction Research
Jackson Laboratory, Bar Harbor, ME (06/2019 – 12/2023)
I designed neural networks for gait detection and pose analysis in mice to support behavioral and addiction research. -
Computational Intern – Small Business Development
Profit Decoder, New York City (06/2023 – 08/2023)
I developed transparent economic data analysis solutions for a start-up supporting small businesses.
Tutoring
I tutor mathematics, computer science, and exam-prep for all ages. I have over 7 years of experience, and like to plan fun, personalized lesson plans focused on problem-solving and holistic learning.
Lessons are available remotely over Zoom, with flexible rates.
Elementary & Middle School
- Math foundations
- Algebra and geometry
- Problem-solving skills
- Introductory programming concepts
High School
- Trigonometry
- Precalculus
- Calculus
- Introductory programming
- PSAT and SAT preparation
- AP Calculus
- AP Statistics
University
- Linear Algebra
- Probability
- Analysis
- Discrete Mathematics
- Data Structures
- Algorithms
- Operating Systems
- Programming coursework support
Interested in working together? Get in touch below.
Contact
Have an idea for an NLP project? Need help with ASR or forced alignment? Interested in tutoring? Reach out at:
📧 mmd9604 [at] nyu [dot] edu