Description

I'm looking for missions that involve:

workflow design
feature finding
statistical analysis
model building
optimisation
data visualisation

I am also available for consultation on ML enhancement on any kind of project. I offer short-length data-science courses for institutions and enterprises, which seek in-house solutions to long-term maintenance of data tasks.

I'm a data scientist with a long term experience in a complex project of 3000 people with several layers of processes, such as the experiments at LHC, CERN. As I was in charge of software development, I worked on all these layers:

data acquisition
distributed computing on large volumes of data
software development coordination
software validation
simulations
detection and clustering algorithms
statistical analysis and presentation of results

With my software and statistical analysis experience, I moved on to projects with more general contexts, and worked on machine-learning specifics for several years. As I also have a background on image processing, computer-vision related projects have been my main competence.

At the same time, I have prepared data-challenge tutorials, and my own python teaching programme. I am very good at teaching, therefore my services would be a optimal for enterprises with existing engineers.

Languages

English
Native or bilingual
French
Fluent
Turkish
Native or bilingual
Persian
Conversational

Workplace preferences

Can work on-site

Paris (up to 50km), Paris (up to 100km)

MIT
Software Developer, Graduate
RESEARCH
September 2006 - January 2009 (2 years and 4 months)
Cambridge, United States
Study of physics heavy-ion collisions with the CMS experiment at CERN
Study and software design for collision characterisation (C++)
Implementation of simulation software interfaces (Fortran, C++)
Design and maintenance of Distributed Computing for Simulations (Shell, Perl, Condor)
Maintenance of data storage (Shell)
Design of testing environment for data processing (Shell)
Calibration of detected raw signals (C++)
Shell C++ Condor Distributed Computing Fortran Perl Statistics
Laboratoire Accélérateur Linéaire
Data Scientist, Post-Doc
RESEARCH
December 2016 - Today (9 years and 6 months)
Orsay
Machine-learning research
Anomaly detection in particle physics data (Python)
Design of workflows for tracking algorithms (Python)
Design of workflows for symbolic regression (Python)
Tests of optimisation methods for symbolic regression (SciPy)
Python Scikit-learn keras
Laboratoire Leprince Ringuet - Ecole Polytechnique
Marie-Curie Fellow
RESEARCH
January 2015 - December 2016 (1 year and 11 months)
Palaiseau
Classification of b-quark jets (C++)
Background subtraction for quark/gluon jet pairs (C++)
Correlations between jets and tracks (C++)
C++ Statistics Marie-Curie