Hi, I'm Khadija Oubaha
AI Engineer
An AI & Computer Science Engineer who works across machine learning research, data pipelines, and production systems, not just proof-of-concept notebooks.
I like taking a problem from an unclear starting point all the way to something real and working.
Casablanca, Morocco
Available Now
Turning Data Into Insights
AI & Machine Learning Engineer
What Drives Me
Languages
Python, SQL, Java
Education
Engineering Degree in AI & Computer Science, ENSAM Casablanca
Skills
Pandas, NumPy, PostgreSQL, MongoDB, Git, GitLab CI/CD
I'm an AI & Computer Science Engineer who builds systems end to end, not just prototypes in a notebook.
My work spans machine learning, data engineering, and production software. I enjoy learning new tools, refining my past projects, and applying rigorous, data-driven thinking to real problems.
Hands-On Experience
Three internships spanning ML research, applied AI products, and enterprise tooling
R&D Engineering Intern UM6P
Designed and benchmarked three deep learning architectures, including a Fourier Neural Operator, as a fast surrogate for a physics solver, raising accuracy from 87.7% to 92.2% via a physics-informed loss.
AI Engineering Intern WELEARN
Designed and delivered two AI systems for the e-learning platform: a Streamlit analytics Dashboard with automated data pipelines and a RAG-based AI Chatbot using hybrid retrieval and Llama 3.
Research Assistant Oracle
Contributed to Oracle's AutoMLx platform through reporting improvements, explainability standardization, metric refinements, and cross-framework regression testing.
Featured Work
A showcase of my recent projects demonstrating expertise in machine learning and data engineering
SecureDoc AI -- Document Anonymization Platform
A 4-service microservices platform that detects and redacts personal data on PDFs using a hybrid regex + BERT-NER engine, deployed on Kubernetes.
Medicinal Plant Identifier
A fully offline mobile app that instantly identifies medicinal plants from a single photo and flags toxicity risks in real time, powered by a fine-tuned InceptionResNetV2 model at 87.37% test accuracy.
Automated Crypto Data Pipeline -- Real-Time BI
A team project delivering an automated Power BI pipeline (CoinGecko API to Google Apps Script to Power BI) tracking market cap, momentum, and dominance in real time.
