Fabio Tesser
Product Tech Designer at Fairmat
Greater Trieste Metropolitan Area Italy
Summary
👤
Senior🎓
Top SchoolFabio Tesser is a Product Tech Designer and AI leader with 14 years of experience building ML-driven simulations and speech technologies from research to product. He has led AI at Fairmat and co-founded Mivoq, applying HMMs, CARTs, PCA, neural networks and deep learning to finance and speech synthesis/recognition problems. His work spans core research (PhD-level) and hands-on engineering, including notable open-source contributions to MaryTTS and transfer-learning improvements in deepset-ai's FARM repository. Fabio combines voice processing expertise—automatic labelling, voice conversion and lip-syncing—with pragmatic production engineering, routinely upgrading legacy systems to modern formats. Based in the Greater Trieste area, he is skilled at turning complex signal-processing research into usable products and libraries. An under-the-radar strength is his history of bridging academic projects and industrial deployments, making him fluent in both rigorous modeling and shipping reliable systems.
14 years of coding experience
22 years of employment as a software developer
Università degli Studi di PadovaPhD Information and Communication Technologies, PhD Information and Communication Technologies at Università di TrentoHigh School Degree Electrotechnics, High School Degree Electrotechnics at Istituto Tecnico Industriale Statale G. Segato Belluno
Github Skills (16)
javas10
pytorch10
transfer-learning10
text-to-speech10
nlp10
language-model10
speech-synthesis10
python10
question-answering10
pre-trained-model10
java10
bert9
roberta8
build-automation8
hmmlearn8
Programming languages (8)
Github contributions (5)
marytts/marytts
Feb 2012 - May 2013
MARY TTS -- an open-source, multilingual text-to-speech synthesis system written in pure java
Role in this project:
Back-end Developer
Back-end DeveloperContributions:118 commits, 3 comments in 1 year 3 months
Contributions summary:Fabio primarily focused on refactoring and upgrading the `Mary TTS` system, specifically related to the voice builder component. Their work involved converting and updating voice files from Mary 4.0 to Mary 5.0 format, including modifications to the `Mary4To5VoiceConverter.java` file. They also addressed issues related to the automatic labelling feature within the `HTKLabeler.java` class. Additionally, the user contributed to the wikidb scripts.
pure-javaspeech-to-textsynthesisspeech-recognitionspeech-synthesis
deepset-ai/FARM
Feb 2020 - May 2021
:house_with_garden: Fast & easy transfer learning for NLP. Harvesting language models for the industry. Focus on Question Answering.
Role in this project:ML Engineer
ML EngineerContributions:6 reviews, 16 commits, 18 PRs in 1 year 3 months
Contributions summary:Fabio primarily contributed to the "farm" repository, which focuses on transfer learning for NLP and question answering. Their commits focused on modifying the language model loading process, including adding an option to manually define the language model class. They addressed issues related to model configuration and the handling of language attributes within different language models. Additionally, they implemented fixes and improvements, such as removing baskets without features, adding warnings, and improving the test setup.
bertxlnet-pytorchnatural-language-processingtransformersharvesting
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.