Zakarya ROUZKI

A Data Scientist.

1. Experience

Bell Canada

Bell logo

Senior Data Scientist

March 2024 - Present

  • Leveraging machine learning algorithms to enhance customer-agent pairing and improve overall customer experience.
  • Focused on optimizing the matching process to ensure that each caller is efficiently connected with the most suitable agent, resulting in increased sales, reduced deactivations, and higher customer satisfaction.
  • The data-driven approach to customer service optimization significantly contributed to the company's revenue generating units (RGU's), while also elevating the level of customer engagement and loyalty.

Data Scientist

November 2022 - August 2024

  • Bell is Canada's largest telecommunications company, providing Mobile phone, TV, high-speed and wireless Internet, and residential Home phone services.

FYCOMPUTIG / IREVOLUTION IRevolution logo

Data Scientist - Full-time

Aug 2020 - Present

Rabat, Rabat-Salé-Kenitra, Morocco

  • Data Scientist contributing to the development of Data Science/AI solutions.

Ziwig Health

  • Collection, processing, and visualization of patient data.
  • Development of ML models to diagnose patients suffering from chronic diseases (Endometriosis) based on features extracted from RNA analysis.
  • Developed an AI assistant to help answer questions about the solution developed.
  • Data creation, adaptation, and augmentation for the AI assistant using different approaches.
  • Deployment of the AI assistant on multiple platforms such as WhatsApp, Messenger, React-Native app.
  • Enriching platform with other production-ready solutions such as Search Engine, Article Recommendation using modern NLP solution: Haystack.
  • Created a disease entity extraction model for both French and English languages (SpaCy/CamemBert).
  • Dashboarding: Creation of a dashboard using ElasticSearch and Kibana, to help follow the platform development: users increases, survey responding, etc.

Attrition Risk Platform

  • Collection, processing, and visualization of employees data.
  • Aggregation of data using different techniques: Sliding Window Aggregation, Last Window Aggregation.
  • Feature selection, transformation, and normalization.
  • Model building using different algorithms: Xgboost, Random Forest, etc.
  • Handling imbalanced dataset using different approaches: Undersampling techniques, SMOTE.
  • Hyperparameter tuning with GridSearch/RandomizedSearch/Bayesian optimization.
  • Evaluation and models comparison with respect to the business use case.
  • Using Model Explainability techniques: Shapley Values to explain model prediction.
  • Model deployment on AWS EC2 instances, Docker, FastAPI.
  • Built a monitoring dashboard to better measure key model performance metrics.

Skills: Machine Learning · Natural Language Processing (NLP) · Amazon EC2

FYCOMPUTIG / IREVOLUTION IRevolution logo

Data Scientist - Internship

Feb 2020 - Aug 2020

Rabat, Rabat-Salé-Kenitra, Morocco

  • Project: Star5 Intelligence
  • An AI SaaS solution based on Natural Language Processing to track and measure consumer perception of brands and products.

Tasks realized:

  • Multi-Label Classification Engine: classify reviews.
  • Keyword Extraction Engine: extract keyword from reviews.
  • Text Similarity Engine: calculate similarity between two segments of text.
  • Text Segmentation Engine: segmenting reviews into sentences.
  • Sentiment Analysis Engine: detecting sentiment of sentences.
  • Text Summarization Engine: summarizing documents and reviews.
  • Named Entity Recognition Engines: building a specific use case for NER.

Deep Learning Architecture:

  • RNN | LSTM | GRU | SEQ2SEQ | Attention Mechanisms | Transformers.

Transformers Models used:

  • BERT | DistillBert | ULMFiT | T5 | CamemBERT.

Cloud Providers:

  • AWS: Amazon SageMaker | AWS Lambda | AWS API Gateway.
  • Azure: Azure VM's.

Other:

  • Word Embeddings: Glove | FastText | Word2vec.
  • Libraries: PyTorch | Tensorflow | Transformers library by Hugging Face | Fastai | NLTK | Spacy | Scikit-Learn.
  • Model deployment: AWS SageMaker EndPoints & API's | Flask APP on a docker container.
  • Version Controlling: Git (GitLab).
  • Other techniques: Data Augmentation, Data Labeling.

Skills: Machine Learning · Natural Language Processing (NLP) · Amazon Web Services (AWS) · AWS SageMaker

2. Education

Abdelmalek Essadi University University logo

Master’s degree, DATA SCIENCE AND BIG DATA
2018 - 2020

Data Science, Big Data Architecture, Big Data Analytics, Data Mining, Machine Learning, NoSQL Databases, Statistics, Cloud Computing & Virtualization, IoT & Artificial Intelligence, Computer Vision, Databases Administration…

Faculty of Science and Technology, Tangier, Morocco (www.fstt.ac.ma)

Sultan Moulay Slimane University University logo

Bachelor’s degree, Computer Science
2017 - 2018
Grade: Quite Good - B+

Java, C, JEE, Conception (UML & Merise), .Net, Oracle, SQL SERVER, Information System, Networking, Web Development, Project management…

Faculty of Science and Technology Béni Mellal, B.P: 523 Béni - Mellal Maroc