Nicolas Brichler
Verified Expert in Engineering
Data Science Developer
Passionate about machine learning and AI, curious and value-driven, Nicolas is a problem solver seeking the use of new technologies to digitize and modernize companies. Genuinely interested, he always makes an effort to learn about the business side of the projects he's working at to see things from the customer's point of view. Nicolas is looking to contribute to NLP projects related to churn prediction, lifetime value, and sensitivity promotion.
Portfolio
Experience
Availability
Preferred Environment
Python 3, Databricks, Google Cloud Platform (GCP)
The most amazing...
...project I've developed is a natural language model used by Solvay to classify and transfer all internal tickets to the most qualified team.
Work Experience
Senior Data Scientist
Grab
- Supported ads and marketing departments by predicting customer demographic attributes.
- Estimated customer price and promotion sensitivity through causal inference by building an S-Learner model predicting rides' booking success.
- Handled ad-hoc data science tasks like ETL pipeline modifications, deployment, retraining, and maintenance of ML models.
Data Scientist
Solvay
- Assisted the time-series analysis process, forecasting customers' orders and prices (ARIMA) and helping chemical engineers run diagnostics and increase production. Used feature selection methods like LASSO and correlation to find relevant indicators.
- Performed R&D leveraged past solubility experiments using a random forest model to predict the most promising leads and help researchers reduce unsuccessful experiments.
- Optimized the supply chain and the constrained resources allocation, allowing optimal merchandise transport from plants to customers.
Experience
Classification of Internal Tickets
The classification model leveraged pre-trained transformer DistilBERT models through transfer learning to achieve high accuracy. The classifier was trained and deployed on GCP to achieve 100% uptime and availability.
Root Cause Analysis for Chemical Plant Underperformance
I extracted four years' worth of sensors data from the plants and created an XGBoost model predicting plant yield with good accuracy, then used model coefficients and Monte Carlo simulations to understand which inputs were most negatively affecting the yield.
Thanks to this analysis, the engineering team was able to design and implement a solution to partially reduce the loss.
Skills
Languages
Python 3, Python, SQL
Libraries/APIs
TensorFlow, PyTorch, PySpark, Scikit-learn, Pandas, NumPy, XGBoost, Spark ML, Keras
Tools
Spark SQL, BigQuery
Paradigms
Data Science
Platforms
Dataiku, Databricks, Google Cloud Platform (GCP), Azure
Other
Machine Learning, Natural Language Processing (NLP), Artificial Intelligence (AI), Algorithms, Time Series, Big Data, Forecasting, Regression, Classification, Text Classification, GPT, Generative Pre-trained Transformers (GPT), Statistics, Modeling, Mathematics, Time Series Analysis, Deep Learning, Root Cause Analysis, Business Analysis, Data Engineering, Sales Forecasting, Cash Flow Forecasting, Big Data Architecture, Probability Theory, Optimization, Graph Theory, Industrial IT, Chemistry, Monte Carlo Simulations, Causal Inference, Neural Networks
Frameworks
Spark
Education
Master's Degree in Statistics
Universite Paris-Sud - Paris, France
Master's Degree in Mathematics and Computer Science
École Polytechnique - Paris, France
Certifications
Data Engineering, Big Data, and Machine Learning on GCP Specialization
Coursera
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring