Career Profile

I am a Computer Scientist specialised in Machine Learning and Signal Processing passionate of research and software development with a previous career as Software Engineer for both real-time embedded systems and distributed systems. I have also acquired a strong knowledge of information theory, control theory, electronics and telecommunication through my bachelor in Computer Engineering and my master in Computer Science. In my free time, I experiment with robotics and electronics or develop open source projects that vary from games to application to improve my workflow.

Experiences

Instrumentation & Control Engineer

2021 - present
ITER, Saint-Paul-lez-Durance (France)

Working on the fast control system of the ITER Electron Cyclotron plant.

R&D Software Engineer

2019 - 2021
CS Group, Toulouse (France)

Technical leader of the Sentinel 2 Toolbox and responsible of the SNAP continuous integration platform, both for the European Space Agency. SNAP expert support for the RUS Copernicus project. Additional support to other CS teams on algorithm and optimisation.

Full-stack Developer

2016
WedTrade, Geneva (CH)

Development of a Business to Business e-commerce website from scratch using the last web technologies.

Real-Time Software Engineer

2014 - 2016
CERN, Geneva (CH)

Responsible of the real-time embedded control software for several accelerator instrumentation and the front-end expert tool to control them. Other side projects included the development of several GNU/Linux driver for custom VME boards, a web server to access to the VME boards using a simple REST API and developing and maintaining of several Java libraries for the team.

Java Software Engineer

2012 - 2014
CERN, Geneva (CH)

Developed a distributed monitoring system for the CERN accelerator infrastructure called “Inspector“ with multi-node servers and graphical client and editor to create the monitor GUI and a simple Graphical Programming Language to perform online analysis of the instrumentation data.

Projects

LSC and DIP - Python implementation of two novel methods to inject image priors into learnable compressive subsampling.
SciDown - Definition of an extension of MarkDown for scientific publication. Implementation of an HTML and LaTeX compiler written in C.
Charter - Simple Domain Specific Language (DSL) and C library to generate SVG plots.
NBodies - Graphical N-Body simulation in Haskell using the Barnes Hut technique.
Orbital Ballistic - Small game simulating an inter-planetary war written in Lua.
OpenGoo - A free clone of “World Of Goo written” in C++ using the Qt framework.

Publications

Injecting Image Priors into Learnable Compressive Subsampling
M. Ferrari, O. Taran, T. Holotyak, K. Egiazarian, and S. Voloshynovskiy
26th European Signal Processing Conference (EUSIPCO), Rome, Italy
Signal sampling with joint training of learnable priors for sampling operator and decoder
M. Ferrari, O. Taran, T. Holotyak, K. Egiazarian, and S. Voloshynovskiy
European Patent 18191494.6

Skills & Proficiency

C/C++

Java

Python

Octave / Matlab

Julia

Lua

Haskell

HTML5 & CSS & JavaScript

PHP & MySql

Swift & Go

VeriLog