Computer science
Hi, I'm Anaïs — a data scientist with a background in business and computer science. I work on solving real-world problems using machine learning, data analysis, and predictive modeling.
I'm currently looking for data science roles where I can apply my skills to meaningful projects and continue growing in the field. Feel free to contact me if you like what you see!
Response time: Usually within 24 hours
Programming languages:
- Python
- HTML/CSS
- SQL
- R
- Javascript
Skills:
Technical
Soft
Technical proficiency:
Data Science
Pandas, NumPy, Scikit-learn, TensorFlow/Keras, PySpark, Matplotlib, Seaborn, Tableau, R Studio
Machine learning
Supervised learning, Deep learning, Object detection, Predictive modeling, Model evaluation & Tuning
Web development
Frontend basics with HTML/CSS & Javascript, UI/UX design, Deployment
Development Tools
Git/Github (Version control, Issue-tracking), Jupyter Notebooks, VS Code, Docker
Projects
From web development to machine learning, I've completed several projects in a freelance, personal and academic setting.
Most are available on Github and some have been deployed to the web. You can filter through them by date, skill and complexity.
Corto Safaris website
Designed a responsive tourism website with Bootstrap, custom jQuery, and PHP backend for secure data processing.
Implemented modern UI elements, form validation and Google Analytics that increased online booking inquiries and strengthened the company's digital presence in East Africa & Europe.
Bank Management System
Built a Python CLI banking system featuring transaction processing and account management. Implemented OOP principles with inheritance for different account types, transaction history tracking, and data persistence through file I/O operations.
Marketing campaign predictor
Developed a machine learning classifier that accurately predicted bank marketing campaign outcomes.
Evaluated multiple classification algorithms using customer demographic to achieve optimal prediction accuracy, providing actionable insights for more targeted marketing approaches.
Predicting steel-plates defects
Built a PySpark ML pipeline for steel-plate defect classification in a Kaggle competition, achieving an AUC-ROC score of 0.86375.
Implemented distributed computing techniques to process industrial sensor data and develop a multi-class model for manufacturing quality control. Ranked in the 65th percentile among 2199 participants.
Prediction, Image processing & Transfer learning
Created a deep CNN project for audio classification and image recognition and leveragin transfer learning. Fine-tuned pre-trained models with hyperparameter optimization to achieve an improvement over baseline performance.
Magpie
Led comprehensive user testing for Magpie, a GIS platform for urban planners, gathering and implementing critical feedback to enhance usability.
Specialized in training object detection model (YOLO) to accurately identify vehicles from satellite and street-level imagery. Refined detection algorithms to improve accuracy across diverse urban environments, building a brand new dataset of parking spaces.
To do list application
Developed a PyQt5 task management application with an intuitive UI and SQLite database backend.
Implemented priority color coding, recurring tasks and a history function. Applied MVC architecture for clean separation of data and presentation layers.
Portfolio website
Built and deployed a custom portfolio website with HTML, Tailwind CSS, Javascript and icons from SVGrepo.
Leveraging Github for project management, issue tracking and deployment to Firebase through Github actions.