Bienvenue sur le profil Malt de Aurelien !

Malt vous permet d'accéder aux meilleurs consultants freelances pour vos projets. Contactez Aurelien pour échanger sur votre besoin ou recherchez d'autres profils sur Malt.

Aurelien Vermylen

Data Scientist

Peut se déplacer à Bruxelles

  • 50.8465
  • 4.3516
  • Tarif indicatif 778€ / jour
  • Expérience 7 ans et +
Proposer un projet La mission ne démarrera que si vous acceptez le devis de Aurelien.

Disponibilité non-confirmée

À temps partiel, 1 jour par semaine

Proposer un projet La mission ne démarrera que si vous acceptez le devis de Aurelien.

Localisation et déplacement

Bruxelles, Belgique
Peut travailler dans vos locaux à
  • Bruxelles et 50km autour



  • Français

    Bilingue ou natif

  • Néerlandais

    Bilingue ou natif

  • Anglais

    Capacité professionnelle complète


Compétences (18)

Aurelien en quelques mots

My mission is to help companies take strategic decisions based on data, mathematical or statistical models.
I have two strong expertise domains:
- Advising on technology-related decisions, design the architecture of data applications, help move to the cloud.
- Analysing and modelling data in order to take strategic decisions.

I'm a deeply technical person, and can actually implement my advice, without the need for other vendors or subcontractors. I have extensive experience in explaining complex IT and financial concepts to non-technical people which makes me also apt to communicate with executive management.




IT Quant

Bruxelles, Belgique

avril 2018 - août 2019 (1 an et 4 mois)

Project within GEM's Front-Office quant team, improving their commodity price simulator:
- Ported large python2 codebases to python3, introducing proper development practices (source control, testing), fixing problems with legacy dependencies.
- Set up APIs for the core C# simulator with pythonnet and Flask to allow real-time querying.
- Set up periodic and on-demand computations of the price simulator using cron and an in-house task launcher, again allowing real-time demands.
- Manage the compex multi-language stack interoperability (python, C#, R)
- Advise on model usage (historical vs. risk-neutral usage, characteristics of the model dynamics, ...)
- Manage the DevOps chain, set up automatic deployment using Byrd.


Arts & artisanat

Digital Web Specialist

New York, États-Unis d'Amérique

août 2018 - août 2018

Short mission which aimed at lazy-loading ads in the news website of artnet to improve viewability performance. Improved DFP (Google Ad Manager) viewability percentages from below 50% to around 70%.




Lille, France

mars 2017 - mars 2018 (1 an)

Helping ClearRoad showcase how it can greatly simplify road infrastructure owners’ and road service providers’ revenue generation with the help of new, cheap technologies.

- Make a POC application on ClearRoad's existing transaction processor in python, with the use of crossbar (, python and Flask.
- Implement ClearRoad’s data collection services using vendor IOT devices that are plugged in the diagnostics port of cars (OBDII) and transform geospatial data into billable miles clearly identified on a specific road. Use of ERP5 (, SlapOS ( the python scientific stack for the geospatial treatment and jIO ( for synchronization between services and hosted a private openstreetmap database (


Banque & assurances

Model Validation Quant

Bruxelles, Belgique

avril 2015 - février 2017 (1 an et 10 mois)

In the Independent Review (Validation) team for the group-wide trading book risks. Responsibilities included:
- Review of the IMM Counterparty Risk methodologies (focus on IR and FX). Periodic Reviews: working on Interest Rate/FX risk factor models and Commodities risk factor models. Closure of regulatory recommendations: IR and FX option pricer reviews (American, CMS, Range Accrual …), back-testing, collateral modelling, etc. Review of methodology changes: e.g. use of expert models, Negative Interest Rates.
- Market Risk methodologies, Proxy Credit Spread methodologies.
- Research topics like SABR variants under negative rates, new bilateral initial margin requirements, coherent pricing and discounting under different collateral setups, CVA and XVA topics.
- Use and implementation of a variety of statistical and applied mathematics techniques: regression, optimization, maximum likelihood, symbolic mathematics software, linear algebra... Always ensuring efficient implementation in the team’s C++ library or in ad-hoc python scripts.

Dexia - Belfius

Banque & assurances

Quantitative Developer

Bruxelles, Belgique

mai 2011 - avril 2015 (3 ans et 11 mois)

Position of Quantitative Developer at the group level Risk Management team. Team focus was non-market VaRs (mostly Credit). Responsibilities included:
- Ecap Value-at-Risk (VaR) modeling: Statistical modeling for calibration of various credit VaR parameters.
- Analytical Ecap approximation: Calibrating analytical formula to estimate a position's capital consumption based on its credit VaR.
- Stress Testing modelling: Development of macro-economic “elasticity” models to forecast default rates under given scenarios.

Frederic Frederic Charlier - ClearRoad


Excellent développeur, travailleur loyal et team player.