Thomas Manzini
PhD Student - ML & Robotics for Disaster Response


I am currently a PhD student with Dr. Robin Murphy at Texas A&M University.

My interests lay the intersection of machine learning and disaster response. Specifically, how to improve robustness and usability of automated systems in extreme environments.

I grew up in Las Vegas and I enjoy running, hiking, spending time outdoors and in my spare time I am also an Instrument Rated Private Pilot, Advanced EMT, and a Firefighter.

I graduated with my Masters from Carnegie Mellon's Language Technologies Insitutute and my Bachelors from Rensselaer Polytechnic Institute. My educational background is in NLP and ML and my graduate research included collaborations with Bosch GmbH, DARPA, and Boeing.

After graduate school I joined Microsoft where I split my time between product focused ML teams and collaborations with groups like the CDC and the WHO in development of machine learning and data management systems to respond to disasters like the COVID-19 pandemic.

Download CV

Recent Publications

Examination and Extension of Strategies for Improving Personalized Language Modeling via Interpolation

Shao, Mantravadi, Manzini, Buendia, Knoertzer, Srinivasan, Quirk

Workshop on Natural Language Interfaces, Association for Computational Linguistics 2020

Seattle, Washington, USA

Paper Presentation

Black is to Criminal as Caucasian is to Police: Detecting and Removing Multiclass Bias in Word Embeddings

Manzini, Lim, Tsvetkov, Black

NAACL-HLT ’19. The Proceedings of the 2019 Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies

Minneapolis, Minnesota, USA

Paper Presentation Repository

Found in Translation: Learning Robust Joint Representations by Cyclic Translations Between Modalities

Pham, Liang, Manzini, Morency, Póczos

AAAI ’19. Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence

Honolulu, Hawaii, USA

Paper Presentation Poster Repository

View All




Honors & Awards

Tutorials & Demos

Community Service