A few more things.
Outside of pipelines and papers, most of what I do still comes back to the same instinct: check whether something is actually precise, or just sounds precise.
Mathematics, mostly
I'm reading for a B.Sc. in Mathematics at Yıldız Technical University in Istanbul (2022–2026). I came to data work from proofs rather than from a computer-science track, and it shows: my first question about any system is usually what the definitions are, and what happens when the real world stops matching them.
That question connects the three things I spend time on — a segmentation study, a Redshift warehouse and a fleet of automation bots. Different tools, same failure mode.
Where curiosity goes
I started in automation — RPA at Latro and Papara, wiring SAP and internal tools together with Robusta and Automation Anywhere. That work taught me to be suspicious of anything that "usually works," because a bot doesn't get to shrug at an edge case.
From there the pull was toward data: pipelines, warehouses, and eventually applied ML through the TÜBİTAK research project. I'm currently a data intern at Getir, and I spend a fair amount of spare time reading around applied machine learning and probability on Kaggle.
Volunteering
Alongside coursework and internships I've coordinated volunteers across two organisations, and still do at Toplum Gönüllüleri Vakfı. Mostly logistics and scheduling rather than anything technical — but a system made of people fails in ways a pipeline never does, and that has been its own education.
Elsewhere
I also spent five weeks in 2023 on Stanford's Code in Place, a live, volunteer-taught introductory CS course — my first formal exposure to teaching-quality software fundamentals outside of self-study.
Say hello
If you're working on something in data or automation — or want to argue about medical-imaging evaluation — I'm at contact@tahaberk.com.