HUSSEINI Hussein

• Building and maintaining resilient batch ETL data pipelines with Spark, Scala and Python to process large amounts of flights data and Machine learning forecasts in Databricks • Orchestrating ETL workflows with Azure Databricks Workflows and Control-M • Optimizing Spark jobs performance and efficiency resulting in 40% decrease in processing time and reduced cost • Supporting different teams in data platform maintenance

• Designing and implementing end-to-end ML solution to automate optimal loyalty programs recommendations resulting in increased merchant revenue • Creating and deploying a REST API (FastApi) and Streamlit applications with Python Kubernetes in AWS • Creating a MLOps pipeline with AIrflow, Docker, MLFlow and Jenkins for model retraining and versioning • Causal analysis and inference to estimate product effect on merchant's revenue and deploying as Spark batch job • Statistical analysis of business data for decision making • Maintenance of batch and streaming ETL pipelines in Scala and Java Spark • Applying best practices from software development, monitoring, testing, security and MLOps

• Creating and training Automatic Speech Recognition systems (Speech-to-text) from scratch for English and French with Kaldi on AWS with GPU setup • Web scraping and training adapted language models (LM) for the medical industry • Optimize language models size and efficiency with pruning • Deploying real-time Chatbot with ASR capability on AWS • Speech processing and ML to detect angry emotions in voice • Audit for Orange Bank chatbot and giving NLU recommendations

• Collecting, labeling datasets and training machine learning and deep learning algorithms to detect clients' emotions based on their reviews • Building a statistical spelling correction tool • Analysis and predictive modeling of accounting data • Time series analysis and forecasting services demand • Weekly results reporting