Gabriele Santin, Developer in Belluno, Province of Belluno, Italy
Gabriele is available for hire
Hire Gabriele

Gabriele Santin

Verified Expert  in Engineering

Machine Learning Developer

Location
Belluno, Province of Belluno, Italy
Toptal Member Since
February 7, 2022

Gabriele is an applied mathematician with ten years of experience in machine learning and computational mathematics, and in their combination to solve challenging problems in simulation and data science. He is currently interested in cracking problems involving structured data and physics-informed models, especially those related to graph machine learning.

Availability

Part-time

Preferred Environment

Python, MATLAB, PyTorch, PyTorch Geometric (PyG), GitHub, Jupyter Notebook

The most amazing...

...project I've developed is a digital contact tracing simulator combining a mathematical forecast model with a spread simulation on a real contact network.

Work Experience

Researcher and Data Scientist

2019 - PRESENT
Bruno Kessler Foundation
  • Developed a digital contact tracing simulator combining a mathematical forecast model with a spread simulation on a real network. The technology has been used in the creation of the work-safety startup, Tech4Safe.
  • Designed and developed an open-source tutorial on PyTorch Geometric, a leading tool for implementing geometric deep learning algorithms. The video tutorials—available at https://github.com/AntonioLonga/PytorchGeometricTutorial—have 50,000 views.
  • Built and executed a federated learning model for a secure training of a distributed anomaly detection system.
Technologies: Deep Learning, Machine Learning, Graphs, Applied Mathematics, Numerical Simulations, Data Science, Scientific Data Analysis, Mathematics, Research

Researcher

2015 - 2019
University of Stuttgart
  • Developed a benchmark to test uncertainty quantification methods for the reliable simulation of geophysical problems. The benchmark has been developed as the result of the collaboration between four research groups across multiple disciplines.
  • Implemented a package for the training of greedy algorithms for kernel-based learning available at https://github.com/GabrieleSantin/VKOGA. The algorithms have been used in several research projects.
  • Designed, analyzed, and implemented a machine learning surrogate to accelerate the flow simulation in a model of the human vascular network. The model achieved a 0.1% relative error of order and a one-million speedup factor over the full system.
Technologies: Applied Mathematics, Algorithms, Machine Learning, Numerical Simulations, Scientific Data Analysis, Mathematics, Research

Federated Anomaly Detection

A federated learning model to train an anomaly detection classifier using distributed and private datasets. The training is designed to permit the collaboration of different agents that share insights to improve their own classifiers while never sharing data.

Digital Contact Tracing

https://github.com/DigitalContactTracing/covid_code
A Python simulator of a digital contact tracing system that combines a mathematical model's numerical simulation with an agent-based spread model on a real contact network. The system allows the user to design, tune, and test intervention scenarios to contain the spread of a virus under several constraints.

Languages

Python

Libraries/APIs

NumPy, PyTorch, Pandas

Tools

MATLAB, GitHub

Paradigms

Data Science

Other

PyTorch Geometric (PyG), Machine Learning, Deep Learning, Numerical Simulations, Scientific Computing, Scientific Data Analysis, Mathematics, Research, Applied Mathematics, Optimization, Data Analysis, Algorithms, Graphs

Platforms

Jupyter Notebook

2012 - 2016

Ph.D. in Computational Mathematics

University of Padova - Padova, Italy

2009 - 2012

Master's Degree in Mathematics

University of Padova - Padova, Italy

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring