Independent work.
Moving from enterprise database development into data engineering. Building pipelines end to end — ingestion, warehousing, dbt modelling and testing — and measuring the German market rather than guessing at it.
Currently building
A pipeline built on a data source I control: synthetic storefront sessions — product views, add-to-cart, checkout, purchase — generated from configurable funnel rates, then processed end to end. The generator and a verifier that checks its output against the configured funnel are done; ingestion and modelling come next.
Projects
I wanted to know which tools German data engineering roles actually ask for, so I measured it instead of guessing. Two public job APIs into Databricks, modelled with dbt, deduplicated across sources and rebuilt every morning. Azure runs about 1.4× AWS — the reverse of the global picture — and roughly two in five postings state an explicit German-language requirement. The exact figures are on the dashboard, and they move.
Telecom call records rated twice — once by a legacy Oracle PL/SQL package, once by a Python port of the same rules — then compared row by row. 20,000 records, zero differences, on Decimal arithmetic so the rounding matches exactly.
A three-broker KRaft cluster in Docker with a Python producer and consumer, so I could watch what partitions, replication and consumer groups actually do instead of reading about them.
A Spring Boot clinic across admin, doctor and patient roles: booking against verified availability, JWT per role, and MySQL for the relational core with MongoDB for prescriptions. Capstone of the IBM Java Developer certificate.
Work
Claims processing on Oracle PL/SQL and C#/.NET at Agito. Insurer and Medula integrations, bulk data operations, query redesign and production incident response.
2024 — 2026BSc final project at Özyeğin University: a blockchain layer letting IoT devices exchange data without exposing device identity or payload contents.
2023 · ThesisBackground
Stack
Links / Contact
atakuruk@gmail.com