Career Profile

With a PhD in AI and an Executive MBA in Digital Entrepreneurship, I offer a unique blend of technical expertise in data science and machine learning engineering, coupled with profound comprehension of business complexities and requirements. My experience spans data science, machine learning, and project management, with a focus on bridging business and technical teams. I thrive on collaborative efforts and have a proactive approach to problem-solving, ensuring project success.

Experiences

Confirmed Data Scientist/ ML Engineer

2023 - Present
IDOS, Paris France

Deeply Project: Online Reputation Management Application for Tourism Professionals

  • Participation in the project scoping phase and contribution to the platform design.
  • Close collaboration with the executive team and business units.
  • Expertise in data science: identifying data sources, defining key metrics for online reputation evaluation, specifying data analysis algorithms, including aspect-based sentiment analysis.
  • Active development of key components of the solution.
  • Tech Stack : Stack technique : Python, SQL, Scrapping, Selenium, NLP, Sentiment Analysis, Hugging Face Transformers, FastAPI, Airflow, Pytest, ELK Stack, GCP, Docker, Kubernetes Git/Gitlab.

Confirmed Data scientist

2021 - 2022
IDOS, Paris, France

DATAMOTOR Project : Sales Forecasting and Pricing Optimization for a Physical Clothing Store

  • Collaborating with client teams to understand and articulate their needs.
  • Collecting and cleaning data from multiple sources to ensure quality and consistency.
  • Designing and developing Machine Learning models to forecast in-store sales and optimize pricing strategies.
  • Deploying ML models into production and monitoring their performance.
  • Tech Stack : Python, SQL, Time Series, Pandas, Scikit-Learn, ARIMA, FB Prophet, XGBoost, Tensorflow Data Validation (TFDV), MLFlow, Airflow, FastAPI, Pytest, ELK Stack, PostgreSQL, GCP, Docker, Kubernetes, Git/Gitlab.

Data Project Manager

2019 - 2020
INRIA Startup Studio, Paris, France

Co-founder of a deep tech startup named “Closanet”

  • Closanet offers advanced recommendations based on image and language processing for online shopping.
  • Designing the IT infrastructure and modeling the data pipeline.
  • Writting functional specifications and product backlogs.
  • Developing an image processing module for fashion trend analysis.
  • Managing a team of three people and completed a POC.
  • Tech Stack : Python, Java, Visual Computing, NLP, Keras, PyTorch, Selenium, Spring Boot, Postgresql, Elasticsearch, Heroku, Git. Agility, Scrum, Kanban, Trello

R&D Artificial Intelligence Engineer

2017 - 2019
INRIA Startup Studio, Paris, France

PICS Project: Personal Information Controller Service

  • Design and development of a personal data search module on the web, powered by a privacy risk score.
  • Completion of the POC, mockup modeling, experimental validation, server configuration, solution deployment, and deliverables documentation.
  • Tech Stack : Java, Python, Scrapping, Selenium, Apache TIKA, ETL, NLP, Privacy, Micro-services, Elasticsearch, Postgresql, NGINX, Balsamiq

Java /JEE Computer Science Engineer

2009 - 2011
Cylande Africa, Tunis, Tunisia
  • Participated in the development of Cyprus ERP, a software designed for the retail industry to better manage retail stores
  • Tech Stack: Java, SQL, JSF, Hibernate, Spring, Oracle, Maven, Git/SVM

Certifications

Executive MBA "Entrepreunership and management in the digital age"

2023
Visiplus Academy

Training program offering specialization in the field of management and digital marketing for promoting businesses in the digital era.

  • Company Audit, Market Analysis, Entrepreneurship & Management, Project Planning & Management, KPIs & Dashboarding.
  • Preparation for agile management of cross-functional teams, development of robust business plans, and planning of viable projects.
  • Final Project: “Proposal for a repositioning strategy for a Startup dedicated to managing the online reputation of tourism professionals for better differentiation in the market.”

Machine Learning on Google Cloud

2024
Google Cloud Training

Real-world experimentation with end-to-end ML on Google Cloud. This specialization have teached me how to:

  • build Vertex AI AutoML models without writing a single line of code;
  • build BigQuery ML models knowing basic SQL;
  • create Vertex AI custom training jobs that I deployed using containers using Docker;
  • use Feature Store for data management and governance;
  • use feature engineering for model improvement and determine the appropriate data preprocessing options for your use case;
  • use Vertex Vizier hyperparameter tuning to incorporate the right mix of parameters that yields accurate, generalized models and knowledge of the theory to solve specific types of ML problems,
  • use Vertex AI prediction and model monitoring services to help manage the performance of our ML models.

Apache Spark (TM) SQL for Data Analysts

2023
Databricks

I learnt how to use Spark SQL and Delta Lake to ingest, transform, and query data to extract valuable insights that can be shared with my team.

Projects

This is some of my personal and professional projects. Some of them are hosted on Github and are publically accessible. **This section is under construction**

Managing the Empirical Hardness of the Ontology Reasoning Using the Predictive Modelling - PhD thesis project, Designing a decision support system based on predictive models that assess the robustness of reasoners and the complexity of ontologies, while also enabling the extraction of complex ontology parts.
Predicting Fuel efficiency using ML and Automate the Experimentation with MLFlow - This project aims to explore and predict the fuel efficiency measured in miles per gallon (MPG) of various automobiles using machine learning techniques. We compared different preprocessing and learning techniques using Scikit-learn, Keras, MLFlow.
Customers Segmentation Classification Challenge - In this project, we explore the data, define the machine learning problem, and provide a solution to classify customers into existing segments. Additionally, we propose a model for re-segmenting customers using K-means. The computed clusters are explained using Shap.
Avocado prices forcasting using Facebook Prophet, univariate vs multivariate solution - In project, we explore the data and develop a predictive model for forecasting future avocado prices in the US using historical time series data. Leveraging the Facebook Prophet algorithm, the project will explore both univariate and multivariate approaches to enhance prediction accuracy with multiple explanatory variables.

Publications

This section showcases a selection of my scientific publications produced during my PhD research. More publications could be found at my ResearchGate page

  • Multi-label Based Learning for Better Multi-criteria Ranking of Ontology Reasoners
  • Nourhene Alaya, Myriam Lamolle, Sadok Ben Yahia
    International Semantic Web Conference, ISWC (1) 2017: 3-19
  • RakSOR: Ranking of Ontology Reasoners Based on Predicted Performances
  • Nourhene Alaya, Myriam Lamolle, Sadok Ben Yahia
    The 28th International Conference on Tools with Artificial Intelligence, ICTAI 2016: 1076-1083
  • Predicting the Empirical Robustness of the Ontology Reasoners based on Machine Learning Techniques
  • Nourhene Alaya, Myriam Lamolle, Sadok Ben Yahia
    The 7th International Conference on Knowledge Engineering and Ontology Developement, KEOD'2015: 61-73

    Skills & Proficiency

    Python, Java

    SQL

    Scikit-learn, Pandas, Numpy, Scipy

    Keras, Tensorflow, Pytorch

    Seaborn, Matplotlib

    FastAPI, Flask, Streamlit

    MLFlow, AirFlow

    JSF, JPA/Hibernate, Spring Boot

    Oracle, PostgSQL, Elasticsearch, MongoDB

    GCP, Vertex AI, BigQuery

    Docker, Kubernetes

    SVN, Git, Maven