I build data pipelines from raw ingestion to analytics-ready tables.

I started in AI, became curious about the data behind the models, and chose data engineering to work on that part. I build batch data pipelines in Python, SQL, and dbt, most recently for Flint, a job search app. Moving from Sana'a to Giza to Haarlem taught me to ask questions before assuming and to adapt quickly, whether I'm facing a new dataset, a new culture, or a new team. I'm now looking for a junior data engineering role where I can keep learning.

• 🌱 Learning Dutch, because I want to feel at home in the Netherlands and work with my colleagues in their language.
• 🎨 Outside of work, I like drawing and learning Turkish.

• Flint is a job search platform for the Dutch job market, built by a team of five (two data engineers, two backend engineers, and one frontend engineer). A daily pipeline collects postings from the Adzuna API and JobSpy, cleans and models them with dbt on Databricks, and publishes analytics-ready tables the backend serves to the app. As one of the two data engineers, I worked on the ingestion from both sources, developed the dbt staging and intermediate models that combine them, and co-built the LLM step that extracts skills, seniority, and contract type from job descriptions. Tools: Python, Pydantic, Airflow, dbt, Databricks, Azure, GitHub Actions Repository · Live demo

• Image Captioning Model Generates descriptions of images by combining a pre-trained InceptionV3 CNN with a Transformer encoder and decoder. My team and I prepared the COCO 2017 data (70,000 captions cleaned, tokenized, and split 80/20 by image) and trained the model with TensorFlow and Keras. Repository