Hello! I'm Zakaria, a data scientist and ML engineer with over 7 years of hands-on experience in building machine learning models, and deploying scalable ML pipelines. I hold degrees from Paris Dauphine University and the National Institute of Statistics and Applied Economics. This portfolio is a curated collection of my projects, showcasing my journey in transforming data into actionable insights. Whether it's building end-to-end MLOps pipelines or diving deep into data science analysis, I'm passionate about leveraging data to solve real-world problems. Let's connect and explore the possibilities!
- Industrializing a Machine Learning model with MLOps - A complete end-to-end ML project.
-
Fine-tuning BERT for Phishing URL Detection - Leveraging Hugging Face transformers for Phishing detection.
-
Building a Financial Analyst chatbot with RAG - Augmenting a Chatbot with Retrieval Augmented Generation (RAG) to analyze LVMH finances.
-
Building a Healthcare Chatbot with Mistral and Gradio UI - Harnessing the power of AI and LLMs to build intelligent, interactive chatbots.
-
Knowledge Distillation for BERT on IMDB Sentiment Analysis - Compressing LLMs to create lightweight models optimized for mobile and edge deployment
-
Byte Pair Encoding (BPE) Tokenizer from Scratch - a key tokenization method widely used in LLMs like ChatGPT to efficiently process text data.
- LSTM-Powered LVMH Stock Price Forecasting: A Multivariate Time Series Approach - Leveraging LSTM networks for accurate time series predictions.
- Gradient Descent from Scratch: A Visual Walkthrough - Understanding and visualizing the core of Gradient Descent algorithm.
- Hypothesis Testing: Student's One-Sample t-test with Scipy - Applying statistical tests for data-driven decision-making.
- Normality Checks Beyond the Histogram - Exploring advanced techniques for assessing data normality.
- Data Drift Detection with Deepchecks - Ensuring model reliability through proactive drift detection.