Lead data scientist

Nitrex Metal inc.

1. Proficient in managing and manipulating large datasets using Spark, an open-source distributed computing system designed for big data processing and analysis.
2. Developing, optimizing and maintaining predictive time series models for example: Long Short-Term Memory (LSTM), AutoRegressive Integrated Moving Average (ARIMA),
3. Convolutional Neural Networks (CNN), and Prophet are all utilized in this project, with Python being the chosen programming language.
4. Lifetime monitoring and anomaly detection.
5. Predictive maintenance.
6. Work in collaboration with project management to deliver effective and timely solutions.
7. Analyze and preprocess raw data: Transform the raw data from various sources (log files, relational database) into an understadable data for the machine learning models. That includes the data cleaning, feature engineering, data imputation, data normalization and many other procedures.
8. Designing scalable architectures and expertise with cloud infrastructure (AWS).
9. Building and shipping products with focus on operationalizing analytics ML solutions to solve real-world problems.

Data scientist

Emovi

1. Data acquisition: Preprocessing, cleaning and verifying the integrity of data used for analysis (Python and Airflow are used to orchestrate the workfow of this stage).
2. Data visualization: Exploring relationships between variables and identifying outliers and patterns using data visualization tools (Python: Pandas, Bokeh and Matplotlib).
3. Analysis of variance (ANOVA): Analyzing the differences among group in a sample and
4. providing a statistical test of whether two or more population means are equal(SPSS).
5. Building the state-of-the-art machine learning algorithms for data classification (Python and TensorFlow).
6. Building and optimizing classifiers using machine learning techniques such as K-NN,
7. Naive Bayes, SVM, Decision Forests and Neural networks (Matlab and Python: Keras, TensorFlow, scikit-learn).
8. Implementing a new statistical approach based on T2 Hotelling hypothesis testing for data classification (Python and Matlab).

Data scientist

Regime

1. Development of an application for multi-language handwriting database acquisition (dotnet, C sharp).
2. Acquisition and publishing of a handwriting database.
3. Optimization of a feature extraction algorithm and beta-elliptic modeling of the hand-writing database (Matlab).
4. Development of machine learning methods for the handwriting database (Matlab and Python).
5. Development of a deep neural network algorithm to improve the handwriting recognition performance (Python).

Institut national de la recherche scientifique

Computer science

Ecole Nationale d'Ingénieurs de Sfax.

Computer science

Faculté des Sciences de Sfax

Computer science and Mathematics