Samuele Piazzetta
Data Scientist
Zurich, Zurich, Switzerland
Summary
🤩
Rockstar🎓
Top SchoolSamuele Piazzetta is a data scientist with seven years of experience, currently applying ML and data-engineering skills at Unit8 in Zurich after earning a Master's in Computer Science from ETH Zurich. He contributes to open-source tooling for time series forecasting—most notably enhancing the popular darts library with transformer multi-series support, parallelization, and robust evaluation features. At Unit8 he mentors junior data scientists, conducts technical interviews, and builds end-to-end solutions from PySpark ingestion pipelines to React frontends for production use. His work spans NLP (Text2SQL, RAG), protein-sequence generative models for pharma, and applied anomaly detection for predictive maintenance, reflecting a blend of research and product delivery. Colleagues value his attention to detail and ability to translate cutting-edge research into reliable, production-ready systems.
7 years of coding experience
Bachelor's degree, Computer Science, 110 e lode, Bachelor's degree, Computer Science, 110 e lode at Università degli Studi di PadovaMaster's degree, Computer Science, Master's degree, Computer Science at Eidgenössische Technische Hochschule Zürich
Github Skills (17)
pytorch10
anomaly-detection10
python10
machine-learning10
time-series10
forecasting10
deeplearning-ai10
deep-learning10
forecast10
numpy9
scikit9
scikit-learn9
eval8
pandas8
evaluation8
Programming languages (4)
Github contributions (5)
unit8co/darts
Apr 2021 - Jan 2023
A python library for user-friendly forecasting and anomaly detection on time series.
Role in this project:
Data Scientist & ML Engineer
Data Scientist & ML EngineerContributions:19 reviews, 100 commits, 15 PRs in 1 year 9 months
Contributions summary:Samuele's primary contributions focus on enhancing the `darts` library, particularly in the domain of time series forecasting and anomaly detection. They implemented multi-TS support for transformers, updated documentation, and refactored transformer classes to improve parallelization. They also worked on integrating metrics and parallelizing grid search, demonstrating a strong understanding of model evaluation and optimization techniques. Further improvements included addressing NBEATS model issues and fixing potential bugs in the codebase, indicating attention to detail and reliability.
forecastingpython-libraryanomalypythontime-series-analysis
piaz97/Assignment_1
Oct 2018 - Oct 2018
Contributions:13 pushes, 5 branches, 6 issues in 1 day
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.