Alexandre Caulier DATA SCIENTIST AT SNIPS 71 boulevard Arago, 75013, Paris, France
(+33) 669918450
|
[email protected]
| Garvys | alexandre-caulier/en
| 22 years old
Professional Experiences Snips
Paris, France
DATA SCIENTIST
August 2017 - Today
• Development of a Speech To Text solution able to run in real time on small devices like a Raspberry Pi 3
Snips
Paris, France
DATA SCIENTIST INTERN
February 2017 - August 2017
• Development of a Speech To Text solution able to run in real time on small devices like a Raspberry Pi 3
Aubay
Boulogne-Billancourt, France
INTERN
June 2016 - July 2016
• Deep study of HandlebarJS • Serveur/client communication optimisation for a web site NodeJS/HandlebarJS/OrbitJS
Airbus Helicopters
La Courneuve, France
INTERN
June 2015 - July 2015
3D Development of a lab. Excel Macro programming.
Education ENSIMAG (École Nationale Supérieure d’Informatique et de Mathématiques Appliquées) ENGINEERING DEGREE
Grenoble, France September 2014 - September 2017
Artificial Intelligence, algorithms complexity, probabilities...
NTNU (Norges teknisk-naturvitenskapelige universitet) 6 MONTHS ERASMUS EXCHANGE
Trondheim, Norway September 2016 - December 2016
Artificial Intelligence Programming. Parralel Computing. Specialisation : Deep Reinforcement Learning
Lycée Albert Schweitzer
Le Raincy, France
PREPARATORY CLASS
September 2012 - June 2014
Specialisation in Mathematics and Physics (MPSI - MP*)
Lycée Saint Louis Ste Clotilde
Le Raincy, France
SCIENTIFICAL BACCALAUREATE, MATHEMATICS AND PHYSICS, GOOD MENTION
September 2005 - June 2012
Specialisation Mathematics
Projects Algorithmic, programming and data science contests (details below) CODINGAME, HACKERRANK, CODECHEF, KAGGLE, GOOGLE HASHCODE, GOOGLE CODE JAM
Grenoble, France September 2014 - Today
Competing in several algorithmic, programming and data science contests, either individually or in a team. Skills : Algorithmic, Data structures, Dynamic Programming, Optimisation, Artificial Intelligence, Data Analysis...
Specialisation Project - Deep Reinforcement Learning - Study of advanced exploration methods REINFORCEMENT LEARNING, ATARI, OPENAI GYM
Trondheim, Norway August 2016 - December 2016
Research on state of the art Deep Reinforcement Learning, which shall lead to implementing some of the most promising ones so as to rank performances and define adequate use for each of them. Implementation of numerous state of the art papers : DQN, Double DQN, Bootstrapped DQN, Prioritized Replay, Value Difference Based Exploration... Deep Reinforcement Learning for Mars Landers Trondheim, Norway REINFORCEMENT LEARNING, NEURAL NETWORKS, TENSORFLOW
September 2016
Creating a deep reinforcement learning artificial intelligence using Tensorflow in Python for Codingame’s Mars Landers. Reading and implementing recent research articles to create a Deep Q-Network with Experience Replay and epsilon-greedy policy.
Contests
Kaggle - Final Data Science Game 2017 - 6th / 20 teams (2nd French team) LIGHTGBM, XGBOOST, TIME SERIES, STACKING
Paris, France September 2017
Final of the Data Science Game 2017, where the 20 best teams of the online qualifier gathered in Paris for the final challenge. The goal was to predict the demand in spare parts for different countries for 3 months. To make these predictions, we were given the historical demand as well as historical data about how these past deliveries: products characteristics, prices... for 5 years. Kaggle - Data Science Game 2017 - 15th / 217 teams Paris, France FACTORIZATION MACHINE, XGBOOST, STACKING
April 2017 - May 2017
Music recommendation challenge. The goal of this challenge is to predict whether the users of the test dataset listened to the first track Flow proposed them or not. We had access to the Deezer API to create new features and develop powerful models.
OCTOBER 18, 2017
CAULIER ALEXANDRE · RÉSUMÉ
1
Codingame - Hypersonic - 8th / 2715 competitors MONTECARLO, OPTIMISATION, EVALUATION
Trondheim, Norway September 2016
Development of an Artificial Intelligence to play the game Bomberman. My solution was based on a Monte Carlo algorithm to explore the game tree and choose the best action within 100 milliseconds. Kaggle - Talking Data - 16th / 1689 teams Paris, France XGBOOST, NEURAL NETWORKS, STACKING
August 2016 - September 2016
Classification contest. We had to predict users demographic characteristics based on their app usage, geolocation, and mobile device properties. Kaggle - Data Science Game 2016 - 26th / 117 teams Paris, France CONVOLUTIONAL NEURAL NETWORKS
June 2016
Image classification contest. We were given a dataset of satellite images of roofs. We had to predict the orientation of the roofs into 4 different categories (North/South, East/West, Flat roof and Others). Kaggle - San Francisco Crime Classification - 10th / 2335 teams Grenoble, France XGBOOST, NEURAL NETWORKS, LINEAR REGRESSION
June 2016
Classification contest. This competition’s dataset provides nearly 12 years of crime reports from across all of San Francisco’s neighborhoods. Given time and location, we had to predict the category of crime that occurred. Regions Job - Battle Dev 2016 - 19th / 1218 competitors Grenoble, France ALGORITHMIC, OPTIMISATION
March 2016
Algorithmic competition that lasted 2 hours. We had to solve 5 problems as quickly possible. Low algorithm complexity was necessary to solve a problem and go to the next one. Codingame - Coders Strike Back - 84th / 2530 competitors Grenoble, France PLANNING, C++, BRUTEFORCE, EVALUATION
February 2016
Development of an Artificial Intelligence able to control two pods on a map with checkpoints. Two players game with collision between pods. Google Hash Code 2016 - 81st / 1054 teams Grenoble, France PLANNING, EVALUATION, C++
February 2016
Google algorithmic competition that lasted four hours. We had a fleet of drones and a list of orders (from different customers). The goal was to optimize the drones path to realize all deliveries as soon as possible .
Associative Experiences Nsigma - Junior Entreprise
Grenoble, France
DEVELOPER
January 2016 - August 2016
Development of a website using Worpress. The site was intended to take order and choose location delivery.
Student Board ENSIMAG
Grenoble, France
EVENT PLANNING
June 2015 - June 2016
Graphic and Communication Manager.
Securimag
Grenoble, France
ENSIMAG COMPUTER SECURITY ASSOCIATION
September 2014 - June 2016
Conferences on computer security. CTF (Capture The Flag).
Ciné’toile
Grenoble, France
BROADCASTING OF A MOVIE EVERY WEDNESDAY
June 2015 - June 2016
Graphic and Communication Manager.
Languages French English
Mother tongue Professional working proficiency - TOEIC : 875 / 990
Skills
Languages Deep Learning Operating Systems Softwares Parallelization
C, C++, Python, Java, Ada, R, Javascript Keras, TensorFlow Windows, Linux, Mac OS Sublime Text, PyCharm, Visual Studio, Eclipse, Microsoft Office, Latex MPI, OpenMP, Pthreads
Interests
Computer science Deep Learning, Reinforcement Learning, Artificial Intelligence, Algorithms Sports Table tennis, Badminton, Hiking, Skiing Personal Cinéma, Series, Reading
OCTOBER 18, 2017
CAULIER ALEXANDRE · RÉSUMÉ
2