Benoit is available for hire
Hire BenoitBenoit Coste
Verified Expert in Engineering
Data Mining Developer
Bordeaux, France
Toptal member since September 17, 2020
Bio
Benoit is a scientist at heart transformed into a software engineer. He worked as an astrophysicist where he did data analysis and data science in order to measure cosmic-ray fluxes. Lately, he's worked as a software engineer for a neuroscience research project where he redesigned a full scientific software stack. His major skills are developing and designing algorithms or improving scientific codes by making them faster or more optimized.
Portfolio
PepsiCo Global
Python, SQL, Pandas, Data Science, Spark, PostgreSQL...
Blue Brain Project
Pandas, NumPy, Python, JSON, YAML, APIs, C++, PyBind11, Plotly, Matplotlib...
Amadeus
NumPy, Python, SQL, C++, AngularJS, MySQL, JavaScript, API Integration...
Experience
- C++ - 11 years
- Pandas - 8 years
- NumPy - 8 years
- Python - 8 years
- Data Mining - 7 years
- API Architecture - 7 years
- Django - 2 years
- SQL - 2 years
Availability
Part-time
Preferred Environment
Emacs
The most amazing...
...thing I've developed is an I/O library for reading and writing neuroscientific morphological data which is now widely used across the Blue Brain Project.
Work Experience
Python Engineer for a Data Science Team
2021 - 2022
PepsiCo Global
- Supervised the refactoring of the codebase. Set up the continuous integration (CI) and the test framework. Optimized code performances, got rid of unused and duplicated code. Shared best practices with other developpers.
- Developped new data analysis pipelines using Spark in order to run clustering algorithms. Set up AWS buckets to cache intermediate computation and store final results.
- Automated a manual workflow into a one click operation. The task went from being one hour long to five minutes long.
- Used Hydra to simplify configuration file management.
- Designed a Python package to interact with the company Jira board (automatically saved command line history, configured files and results into their respective Jira ticket).
Technologies: Python, SQL, Pandas, Data Science, Spark, PostgreSQL, Continuous Integration (CI), Pytest, Data Engineering
Senior Scientific Developer
2017 - 2021
Blue Brain Project
- Developed an open-sourced Python, C++, and I/O library for reading and writing neuron morphology files.
- Supported, modernized, and enriched the morphology stack (around 20 libraries) of the Blue Brain Project.
- Interacted with and managed scientists to make them develop their code according to the project standard. Evangelized them to use our team's recommended software.
- Maintained a PostgreSQL database to keep track of the provenance of each newly generated scientific artifact.
- Developed and maintained our team's continuous integration environment (Travis CI, Github, and Jenkins). Published our open-source software to PyPI.
- Tasked with producing standard compliance Python wheels and Python projects written in C++.
- Developed and maintained an internal website (Django and Vue) connected to our Python internal packages for scientists.
Technologies: Pandas, NumPy, Python, JSON, YAML, APIs, C++, PyBind11, Plotly, Matplotlib, SciPy, PostgreSQL, GitHub, GitHub API, PyPI, Django, Vue 2, JavaScript, Analytics, Data Analytics, Data Visualization, API Integration, Web Development, REST APIs, Open Source, Data Engineering
Python Back-end Developer
2016 - 2017
Amadeus
- Developed optimization algorithms to reseat flight passengers in case of disruptions.
- Maintained and improved the MySQL database storing the passengers in each flight.
- Did on-shift duties to fix bugs found by client airlines.
- Contributed to the front-end development of the website (Angular).
Technologies: NumPy, Python, SQL, C++, AngularJS, MySQL, JavaScript, API Integration, Web Development, REST APIs, Data Engineering
Post-Doctoral Researcher
2014 - 2016
AMS
- Analyzed data to perform the measurement of the cosmic-ray deuterium flux. Used custom statistical Bayesian techniques and machine learning algorithms, including neural networks and boosted decision trees.
- Performed particle physics simulation to improve the understanding of the AMS detector behavior.
- Performed on-shift duties to ensure the communication between the detector HQ and NASA HQ.
- Sampled probability distributions using Markov Chain Monte Carlo techniques.
Technologies: BigQuery, ROOT, NumPy, Pandas, Google BigQuery, Data Mining, Machine Learning, Neural Networks, Gradient Boosted Trees, Bayesian Statistics, Markov Chain Monte Carlo (MCMC) Algorithms, Analytics, Data Analytics, Data Visualization, Data Engineering
Founder and Full-stack Developer for a Startup
2014 - 2015
Croustie
- Launched a mobile app and website to simplify grocery shopping. Set it up so users would pick recipes from a list and get all the necessary ingredients added in the cart of the retailer of their choice.
- Designed the website from scratch. Hosted using Django and connected to a database that would serve all the recipes.
- Scrapped thousands of recipes to get ingredient lists and quantities.
Technologies: Web Scraping, AngularJS, SQLite, Android, Beautiful Soup, Web Development, REST APIs, Data Engineering
R&D Software Engineer
2013 - 2014
Optopartner
- Maintained and developed embedded communication software.
- Explored bugs related to bus communication issues.
- Liaised between the engineer working on the server unit and those on the embedded device.
Technologies: Embedded C, Boost, C++, CAN Bus, API Integration, Data Engineering
Study and Phenomenology of Cosmic-Ray Fluxes
2009 - 2012
Joseph Fourier University
- Modeled a physics simulation of a cosmic ray detector in order to improve the understanding of the detector behavior.
- Performed data analysis in order to measure cosmic ray fluxes. Used Bayesian inference to compute the propagation of statistical uncertainties.
- Implemented a Markov Chain Monte Carlo algorithm to compute the probability density functions of galactic cosmic ray propagation parameters. Published a paper about it.
Technologies: Data Mining, Data Science, ROOT, C++, Markov Chain Monte Carlo (MCMC) Algorithms, Analytics, Data Analytics, Data Visualization, Data Engineering
Experience
MorphIO
https://github.com/bluebrain/MorphIOAn I/O library for reading and writing morphological files.
Neuron morphologies are available in many different formats. This tool aims at providing a unified I/O layer to manipulate morphological data.
It is now the official tool of the Blue Brain Project to interact with the data and so, is used by many higher-level libraries within the project.
In order to reach both users from the Python and the C++ world and also to improve the performances, the library is coded in C++ but provides Python bindings.
Neuron morphologies are available in many different formats. This tool aims at providing a unified I/O layer to manipulate morphological data.
It is now the official tool of the Blue Brain Project to interact with the data and so, is used by many higher-level libraries within the project.
In order to reach both users from the Python and the C++ world and also to improve the performances, the library is coded in C++ but provides Python bindings.
SyncOrg
https://f-droid.org/fr/packages/com.coste.syncorg/An Android note-taking app that supports Emacs OrgMode.
• Modern design using Android’s Material Design guidelines
• Automatic Git synchronization (via SSH and HTTP). Access to the raw files in case of conflict to solve.
• Independent of Emacs: it is no longer required to use it in conjunction with Emacs.
• Agenda view and Todo list view generated from the org files.
• Folding of items and sub-items.
• Share files button
• Modern design using Android’s Material Design guidelines
• Automatic Git synchronization (via SSH and HTTP). Access to the raw files in case of conflict to solve.
• Independent of Emacs: it is no longer required to use it in conjunction with Emacs.
• Agenda view and Todo list view generated from the org files.
• Folding of items and sub-items.
• Share files button
Scientific Publication: A Markov Chain Monte Carlo Technique to Constrain Cosmic Ray Parameters
This publication is about how to constrain a model of galactic cosmic ray propagation using a Markov Chain Monte Carlo technique.
It also includes a work of compilation of critical data (nuclear cross-section) until then scattered across various and sometimes ancient scientific papers.
It also includes a work of compilation of critical data (nuclear cross-section) until then scattered across various and sometimes ancient scientific papers.
Back End of Amadeus Product: Optimized Passenger Recovery
https://amadeus.com/en/portfolio/airlines/passenger-recoveryI was part of the back-end team that implemented the prototype of this project.
The prototype was then successfully pushed into production for one test airline.
A wide range of airline companies now uses the product.
I mostly worked on implementing the Restful API and wrote some of the front-end parts using Angular.
The prototype was then successfully pushed into production for one test airline.
A wide range of airline companies now uses the product.
I mostly worked on implementing the Restful API and wrote some of the front-end parts using Angular.
Education
2009 - 2012
PhD in Cosmic Ray Physics
Joseph Fourier University - Grenoble, France
2008 - 2009
Master's Degree in Physics
Joseph Fourier University - Grenoble, France
2006 - 2009
Master's Degree in Physics
Grenoble INP - Grenoble, France
Skills
Libraries/APIs
NumPy, Pandas, Matplotlib, SciPy, GitHub API, REST APIs, Vue 2, Beautiful Soup
Tools
Plotly, GitHub, PyPI, BigQuery, Emacs, Pytest
Languages
Python, C++, Python 3, SQL, Lisp, Embedded C, Java, YAML, JavaScript, Clojure
Paradigms
API Architecture, REST, Continuous Integration (CI)
Storage
MySQL, PostgreSQL, JSON, SQLite
Frameworks
Django, Boost, AngularJS, ClojureScript, Spark
Platforms
Django CMS, Android
Other
Data Mining, APIs, Scientific Computing, Scientific Data Analysis, Analytics, Data Analytics, Data Engineering, Data Visualization, API Integration, Open Source, ROOT, Data Science, PyBind11, CAN Bus, Google BigQuery, Machine Learning, Neural Networks, Gradient Boosted Trees, Bayesian Statistics, Markov Chain Monte Carlo (MCMC) Algorithms, Web Scraping, Web Development
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