Nicolas Debaene, Developer in Paris, France
Nicolas is available for hire
Hire Nicolas

Nicolas Debaene

Verified Expert  in Engineering

Machine Learning Developer

Location
Paris, France
Toptal Member Since
April 1, 2020

With a master of science degree in engineering and a specialization in data science from Telecom ParisTech, Nicolas has been solving concrete business cases for five years using machine learning and data mining. His experience in the eCommerce, banking, and consulting sectors gives him a very strong technology portfolio and a deep understanding of different business types.

Portfolio

Debaene Analytics
Python, TensorFlow, Pandas, NumPy, Backtesting Trading Strategies...
Common Metal LLC
Data Science, Python, Blockchain, Generative Pre-trained Transformers (GPT)...
Debaene Analytics
Python 3, PyTorch, Solidity, Blockchain, Recommendation Systems, Forecasting...

Experience

Availability

Part-time

Preferred Environment

PyCharm, Jupyter Notebook, SQL, Ubuntu, Windows, Python

The most amazing...

...data science solution that I've built was a credit scoring model that decreases default by €10 million per year for a large French eCommerce company.

Work Experience

Data Scientist | Algorithmic Trading

2022 - PRESENT
Debaene Analytics
  • Developed a trading model with TensorFlow to predict the fluctuations of ETH/USD and BTC/USD perpetual contracts.
  • Built a backtesting framework to test the strategy.
  • Created an end-to-end trading and investment pipeline to run the strategy in production.
Technologies: Python, TensorFlow, Pandas, NumPy, Backtesting Trading Strategies, Options Trading, Data Engineering, Unit Testing, Time Series, Amazon Web Services (AWS), Quantitative Analysis, Predictive Modeling, XGBoost, Financial Forecasting, Machine Learning Operations (MLOps)

Data Scientist with Blockchain Experience for an Investment Company

2022 - 2023
Common Metal LLC
  • Designed a POC to develop a recruiting tool based on Machine Learning.
  • Built a Python tool to extract skills from any resume (NLP).
  • Developed an attrition model to predict employee attrition.
Technologies: Data Science, Python, Blockchain, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Predictive Modeling, XGBoost, Matplotlib

Data Scientist for Algorithmic Trading | Blockchain Developer

2021 - 2022
Debaene Analytics
  • Developed an algorithm (off-chain with Python) to find arbitrage opportunities on Polygon and Binance Smart Chain (triangular arbitrage).
  • Built a smart contract using Solidity to make a gain from those opportunities by doing triangular transactions, and using flash loans when needed.
  • Created a trading algorithm operating on Avalanche decentralized exchanges—a regression model predicting the price of a pair of tokens using blockchain events data and technical analysis.
  • Developed an NFT recommendation engine scrapping data from OpenSea and LooksRare and forecasting the estimated floor price of any NFT collection in the following weeks.
Technologies: Python 3, PyTorch, Solidity, Blockchain, Recommendation Systems, Forecasting, Backtesting Trading Strategies, Code Review, Quantitative Modeling, Web3.js, JavaScript, TensorFlow, Ethereum Smart Contracts, Non-fungible Tokens (NFT), Smart Contracts

Credit Score Modeling

2021 - 2021
PJ Lhuillier Group of Companies - Main
  • Designed and developed an accurate, useful, and stable credit risk model to score customers with existing loan histories on their ability to repay future loans.
  • Created a second model to score prospects with no historical loan data.
  • Developed an optimal credit granting strategy balancing default and acceptance rate depending on the company's tolerated default rate.
Technologies: Data Science, Credit Scores, Modeling, Data Analysis, Code Review, Data Mining, Matplotlib, IPython, Predictive Modeling, XGBoost, Machine Learning Operations (MLOps)

Data Scientist (NLP)

2020 - 2021
Rappi
  • Developed a product categorization model (NLP) from scratch using machine learning and Python to automatically classify products from the eCommerce application thanks to their description and title.
  • Assigned the model to automatically classify four million products sold in the company's eCommerce application in eight different countries in Latin America in 400 different categories, from fresh food, like fruits or meats, to beauty products.
  • Productionized the model that runs automatically when onboarding new products.
  • Developed a matching tool using natural language processing to match new onboarding products to those already existing in the catalog of the eCommerce application, using standard information from products already verified from the catalog.
Technologies: eCommerce, Data Visualization, Jupyter, NumPy, Pandas, Data Analytics, Docker, Scikit-learn, Git, Text Classification, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), SQL, Machine Learning, Python, Data Science, Artificial Intelligence (AI), Code Review, Interviewing, Technical Hiring, Data Mining, Matplotlib, Predictive Modeling, XGBoost, Machine Learning Operations (MLOps)

Data Scientist

2019 - 2020
BNP Paribas
  • Provided artificial intelligence tools to bankers working in BNP Paribas Corporate and Institutional Banking.
  • Detected potential churners by predicting the evolution of sales of Forex products (time series forecasting methods like ARIMA and an autoencoder neural network).
  • Participated in the development of a CV screening tool used by the Human Resources area. Developed an algorithm to find the best path to read a CV (traveling salesman problem: intuitive for humans but very hard to solve for a machine.).
  • Collaborated in developing an optical character recognition tool to extract information from invoices in image format (regional neural network, recurrent neural network).
  • Developed a text classification tool for a CRM system to score descriptions of meetings between a banker and a bank client (a high score means the description is explicit and informative.).
  • Participated in the migration of ETL pipelines (used to create KPI dashboards for the bank) from Alteryx to Spark.
Technologies: Docker, Text Classification, Forecasting, Recommendation Systems, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), Data Visualization, Scrum, Agile Software Development, PySpark, Tableau, Spark, Jupyter, Computer Vision, Data Science, Deep Learning, Machine Learning, Data Analytics, Data Analysis, Artificial Intelligence (AI), Git, Linux, Keras, PyTorch, NumPy, Pandas, Scikit-learn, SQL, Python, Generative Adversarial Networks (GANs), Code Review, Interviewing, Data Mining, Matplotlib, IPython, OCR, Time Series, Predictive Modeling, XGBoost, Financial Forecasting, Document Processing, Machine Learning Operations (MLOps)

Data Scientist

2017 - 2018
CDiscount (eCommerce)
  • Designed and developed an accurate, useful, and stable credit risk model for payment in four installments on the company's selling platform.
  • Initiated the production of new, improved credit scoring models every three months.
  • Optimized credit scoring models to score in real-time every customer buying on the website with one million visits per day, implying building a very fast computing model.
  • Integrated new data sources and solutions into credit risk strategies.
  • Interpreted data science models and gave business recommendations to senior leaders.
Technologies: Docker, eCommerce, Credit Risk, Data Visualization, Scrum, Agile Software Development, TensorFlow, Jupyter, Data Science, Deep Learning, Machine Learning, Data Analytics, Data Analysis, Artificial Intelligence (AI), Git, Linux, Keras, PyTorch, NumPy, Pandas, Scikit-learn, SQL, Python, A/B Testing, Data Mining, Time Series, XGBoost, Machine Learning Operations (MLOps)

Junior Data Scientist

2017 - 2017
Wavestone
  • Created a predictive maintenance model for a transport company using machine learning (data cleaning, gradient boosting trees, and detecting very rare events).
  • Created a benchmark of text classification methods with the purpose to develop a chatbot for the company.
  • Crunched and analyzed data for a French national social organization to detect loss of money in allowances distribution.
Technologies: TensorFlow, PyTorch, Jupyter, Data Science, Deep Learning, Machine Learning, Data Analytics, Artificial Intelligence (AI), Git, Linux, Keras, NumPy, Pandas, Scikit-learn, Python

Customer Churn Detection

While working at BNP Paribas, I developed a statistical and machine learning model to detect potential churners within clients of the bank. I created a method based on time series forecasting to predict the amount of forex transfer each client would request in the next year. The bankers could then take steps to retain the client. I developed a statistical time series forecasting (ARIMA model) and used deep learning auto-regressive models to achieve better performance.

Data Scientist

http://www.rappi.com
• Developed a product categorization model (NLP) from scratch using machine learning and Python to automatically classify products from the eCommerce application thanks to their description and title.
• Assigned the model to automatically classify four million products sold in the company's eCommerce application in eight different countries in Latin America in 400 different categories (from fresh food, like fruits or meats, to beauty products).
• Productionized the model that runs automatically when onboarding new products.
• Developed a matching tool using natural language processing to match new onboarding products to products already existing in the catalog of the eCommerce application (to use standard information from products already verified from the catalog).

Credit Score Modeling

• Designed and developed an accurate, useful, and stable credit risk model to score customers with existing loan histories on their ability to repay future loans
• Created a second model to score prospects with no historical loan data
• Developed an optimal credit granting strategy balancing default and acceptance rate depending on the company's tolerated default rate
2015 - 2017

Master's Degree in Computer Science

Telecom ParisTech - Paris, France

2012 - 2015

Bachelor's Degree in Mathematics

Telecom ParisTech - Paris, France

Libraries/APIs

Matplotlib, Scikit-learn, Pandas, NumPy, TensorFlow, XGBoost, Keras, PyTorch, Web3.js, PySpark

Tools

IPython, Git, Jupyter, Tableau, PyCharm, Apache Airflow

Languages

Python, SQL, Python 3, Snowflake, Solidity, JavaScript, R

Paradigms

Anomaly Detection, Data Science, ETL, Agile Software Development, Unit Testing, Scrum

Platforms

Blockchain, Amazon Web Services (AWS), Ubuntu, Jupyter Notebook, Docker, Linux, Windows

Frameworks

Spark

Other

Artificial Intelligence (AI), Data Analysis, Data Analytics, Data Visualization, Time Series, Machine Learning, Data Mining, Deep Learning, Credit Risk, Natural Language Processing (NLP), Computer Vision, eCommerce, Credit Scores, Modeling, Backtesting Trading Strategies, Linear Algebra, Algorithms, Code Review, Non-fungible Tokens (NFT), Quantitative Modeling, Options Trading, Financial Forecasting, Data Engineering, A/B Testing, Predictive Modeling, Machine Learning Operations (MLOps), Generative Pre-trained Transformers (GPT), OCR, Document Processing, Generative Adversarial Networks (GANs), Recommendation Systems, Forecasting, Bayesian Statistics, Interviewing, Technical Hiring, Smart Contracts, Ethereum Smart Contracts, Quantitative Analysis, Text Classification, Statistics, Software, Physics, Big Data

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