Data science, AI & full stack
ML, data engineering & software
DataCamp · 365 Data Science · IBM
ENSA Khouribga · 2026
Clear value for data-driven teams — from raw data to production-ready AI systems
From problem framing to production deployment — ML pipelines, NLP engines, Computer Vision models. I build systems that deliver measurable results, validated on drawings the models have never seen (duct detection ROC-AUC 0.948, discipline accuracy 92.4%).
Streaming architectures with Kafka & Spark, Big Data processing with Hadoop & Hive. I turn raw data at scale into actionable intelligence — in real time.
BI pipelines, OLAP cubes, Power BI dashboards, and desktop/web interfaces for decision support. I bridge the gap between data and business decisions.
My professional journey in data engineering
@ BIM Pioneers, Casablanca
Designed and built a 3-tier AI platform automating 2D MEP CAD-to-3D BIM conversion, from an AutoCAD plugin to a Revit integration.
@ Maghreb Steel, Casablanca
Built a Stacking ensemble model to predict the steel scrap mass (kg) of each slab from 90k+ rows of production data, delivered as a PyQt5 decision-support tool.
@ Tsunami IT, Remote (Canada)
Built, in a remote team of six students, a real-time multiplayer Scrabble app with Angular and ASP.NET Core, deployed on Azure using Agile workflows.
A selection of data engineering and ML projects showcasing end-to-end solutions
Real-time alert system over 150 weather stations: Kafka and Spark streaming, a GRU forecast (R²=0.85) and XGBoost event detection (F1-Macro=0.67), every alert validated or rejected by an ontology. Angular/FastAPI dashboard.
Comparative study of U-Net, SegNet and PSPNet for face parsing on LaPa (11 classes). SegNet and U-Net are on par (mIoU ≈ 0.75), PSPNet trails at 0.62. Live demo available.
Random Forest model predicting AFCON 2025 match scores from FIFA ranking, recent form and squad value, then a Monte Carlo simulation of the whole tournament (up to 10,000 runs) in a Streamlit app. Morocco comes out as favourite at 29%.
Benchmark of 10 classifiers to anticipate machine failures from sensor data (Random Forest best, 97.9% accuracy on the evaluation split) and a 3-page Power BI dashboard for maintenance teams.
NLP model that scores the compatibility of a CV and a job offer from 0 to 10 (spaCy, TF-IDF, Random Forest and a Keras network, R² 0.83), served by a Flask API and used by the CODEX site to recommend internships.
No-code pipeline built with n8n: a CV submitted as a PDF through a web form is analysed by Google Gemini into structured JSON, stored in Google Sheets and summarised for the recruiter on Telegram.
In about 24 hours, a team of 4 built a course-management platform (Django REST + JWT + MySQL, React) and a K-Means course recommender on an academic dataset, served by a Flask API.
E-commerce platform (React + TypeScript, Spring Boot, MySQL) built around a real quality strategy: 65 backend tests, 46 Selenium end-to-end tests, a JMeter load test, GitLab CI and Docker Compose. Team of 2.
Official website of the CODEX student club: React and Spring Boot (JWT), an admin dashboard, automatic CV parsing and internship recommendations scored by a Flask machine-learning engine. Team of 4.
Three-node Hadoop 3.3.6 cluster on Ubuntu VMs running Python MapReduce jobs over 32 million MovieLens ratings, with two Power BI dashboards.
After finding that Nova-LXD (OpenStack) had been abandoned since 2019, the team pivoted and built a 4-node MicroCloud cluster (LXD, Ceph, OVN) monitored with Prometheus and Grafana.
First-year C console application managing a library's members and books: structures, dynamic allocation, text files and lending rules (3 books max, no duplicates, unique codes).
Technologies and tools I work with daily
Industry-recognized credentials validating expertise in data science and engineering
DataCamp
DataCamp
DataCamp
365 Data Science
DataCamp
IBM Coursera
My academic journey and certifications
ENSA Khouribga
Graduated with Highest Honors (Mention Très Bien). Specializing in Big Data, Machine Learning, and AI. 2 years of preparatory classes followed by a 3-year engineering cycle.
ALX Software Engineering Program
Intensive 12-month program. Specialized in Backend Development, mastering algorithms, system design, and cloud deployment.
Lycée Khadija Oum Al Mouminine
Graduated with Honors (Mention Bien). Focused on advanced Algebra and Analysis, establishing a strong foundation for Data Science.
What I'm studying right now, beyond my day-to-day work
DeepLearning.AI course on Coursera: retrieval, vector search and grounding LLM answers in your own documents.
DeepLearning.AI course on Coursera: how LLMs work, fine-tuning, evaluation and building applications on top of them.
Taking models from notebook to production: packaging, versioning, deployment and automation.
Contributing to the data science and tech community
Enactus ENSA Khouribga
Led digital communication strategies and managed social media campaigns to promote social entrepreneurship projects.
Dataverse Event
Coordinated logistics and planning for the Dataverse event, fostering a community of data enthusiasts.
Enactus ENSA Khouribga
Contributed to various social impact projects, developing teamwork and project management skills.
Hackathons & Open Competitions
Participated in technical challenges to sharpen problem-solving skills under pressure.
Communication abilities and professional traits that complement my technical expertise
Freshly graduated Data & AI Engineer actively seeking my first professional role. Specialized in Machine Learning, NLP, Computer Vision, and Big Data pipelines. Let's build something great together!
