Experience

Laëtis

• France
• Technology, Information and Internet
• 1 - 100 Employee

• Web Developer

  • Jan 2020 - Present

  I build web apps with Django (Python) and Vue.js. I build web apps with Django (Python) and Vue.js.



CNRS

• France
• Research Services
• 700 & Above Employee

• Data Scientist and Web Developer for AI interpretability

  • Jun 2019 - Dec 2019

  I built a Python package and a web application with Django and React to visualize how a black-box machine-learning model behaves and make sure it doesn’t discriminate people based on their gender, skin... Python package: https://github.com/XAI-ANITI/ethik Web app: https://github.com/XAI-ANITI/ethik-ui I worked with Jean-Michel Loubes, Laurent Risser and Max Halford. I built a Python package and a web application with Django and React to visualize how a black-box machine-learning model behaves and make sure it doesn’t discriminate people based on their gender, skin... Python package: https://github.com/XAI-ANITI/ethik Web app: https://github.com/XAI-ANITI/ethik-ui I worked with Jean-Michel Loubes, Laurent Risser and Max Halford.



Simudyne

• United Kingdom
• Software Development
• 1 - 100 Employee

• Data Scientist and Python Developer for Simulation Models

  • Sep 2018 - Jan 2019

  I develop Python tools and visualisations to validate and analyse simulation models, including agent-based models. It includes:* Time series analysis (e.g. time series clustering)* Sensitivity analysis (e.g. Sobol analysis)* Statistical tests (e.g. Kolmogorov-Smirnov test)* Bayesian statistics (e.g. Black-Box Variational Inference, Gibbs sampling)The R&D is done with the Python scientific ecosystem (numpy, pandas...) and involves distributed programming with Dask.The fully-fledge tools are delivered as a Python library I built from scratch. With this library, the user can interact with the SDK (in Java) using the REST API, run models and then read and query produced data to analyse it and visualise it. Show less

• R&D Intern in Data Science for Agent-Based Modelling

  • Mar 2018 - Aug 2018

  I work on statistical techniques and visualizations to validate and explore agent-based models (ABM), especially in finance and economics, that is understanding the influence of the parameters on the outputs and finding parameter combinations leading to interesting behaviors. I try to fill the gap between research and private sector, and I focus on time series, machine learning techniques relevant to them, sensitivity analysis and space exploration (Design Of Experiment, optimization algorithms...). I work with Python.Examples of resources:* Facilitating Parameter Estimation and Sensitivity Analysis of Agent-Based Models: http://jasss.soc.surrey.ac.uk/17/3/11.html* Genetic Algorithms for the Exploration of Parameter Spaces in Agent-Based Models: http://forrest.stonedahl.com/thesis/forrest_stonedahl_thesis.pdf* Validation of Agent-Based Models in Economics and Finance: http://www.isigrowth.eu/wp-content/uploads/2017/10/working_paper_2017_25.pdf Show less



University of Ottawa

• Canada
• Higher Education
• 700 & Above Employee

• Web developer - Open source project IEML

  • Jun 2017 - Aug 2017

  IEML is an artificial language designed by the philosopher Pierre Levy and in which syntax and semantic are related. Just by looking at the characters of two words, it is possible to determine if they have a similar meaning, if one is a specialization of the other, how far they are from each other in the semantic space, etc. During this internship, I developped in a team of four a web application called Intlekt which takes advantage of IEML for data curation. The user can load a collection of documents (articles, videos, images, etc.), describe them with keywords in natural languages (like French or English) then visualize them semantically: are they related to the same topic? what subjects are not mentionned in the collection? etc. I built a REST API with the framework Django (Python) and used MongoDB to store data: https://github.com/IEMLdev/django-intlekt (Vayel). I also participated to the front-end development: https://github.com/IEMLdev/Intlekt-editor. Show less



Nsigma Junior-Entreprise

• France
• IT Services and IT Consulting
• 1 - 100 Employee

• Programmer

  • Sep 2015 - Mar 2017

  I worked as a programmer for the Junior Company Nsigma. Through this role, I was subject to real-world constraints and I learnt to follow specifications and respect deadlines.

• Project Manager

  • Nov 2015 - Feb 2017

  I received projects from clients and helped them define their needs. I wrote specifications and gave estimates of costs. Once the contract was approved, I managed the programmers to make the development progress correctly.I also developped intern tools to automate the processes. One of them aimed at making the redaction of administrative documents faster.

• Responsible for Human Resources

  • Aug 2016 - Dec 2016

  I was in charge of finding programmers among students to implement the products our clients had ordered. To do so, I published recruitment ads and collected data about students' skills and availability to be able to determine the best candidates.I also selected and trained my successor. His formation consisted in exercices and debates around the role of Human Resources Director.