personal site · about

hello

i'm tunç. ai engineer at coop, switzerland's largest retailer. i build production ai by day, small saas at night, and write about everything in between.

this page is the long version of who i am and what i'm chasing. the rest of the site is where i actually work through it.


§ background

grew up in basel. my family is originally from the dersim region — zazaki is the language spoken there. at home i grew up speaking turkish, not zazaki, unfortunately — which is part of why keeping it alive matters to me.

studied wirtschaftswissenschaften at the university of basel, then spent most of my twenties in startup land — first a co-founded venture, then a small software shop building for swiss SMEs for about seven years. that's where i learned to code: it started as a necessity and turned into the part of the work i actually loved.

eventually wanted a change of rhythm — less self-employment, more "what does a real organisation look like from inside." back to basel for a master's in business & technology (blockchain, ai & optimisation, databases — a lot of hard problems), graduated in the top ten of the program, then joined a corporate trainee programme. landed in coop's ai / ml analytics team. still here, still enjoying it more than i expected.

the small saas on the side never stopped — mostly built with kaan sentürk. language-islands.com (web, ios, android) and therapiebericht.ch fall in that bucket. zazaki-islands.com is the outlier: a non-profit i run on my own, to keep an endangered language alive by giving its native speakers a place to translate, record, and pass it down.


§ thinking

the through-line, for as long as i can remember:

  1. 01

    understanding over shortcuts

    i'd rather spend three evenings figuring out how something actually works than ship a half-understood version of it. it costs me time. it's also the only mode i fully trust.

  2. 02

    learning is the default

    books, courses, papers, things i can pull apart. when there's something i don't understand, that becomes the priority. work fits around it, not the other way.

  3. 03

    writing to actually understand

    if i can't explain something clearly, i don't know it yet. that's why this site exists — not as a portfolio, but as the place where i check whether i've really understood it.


§ reach

email is the best way to reach me. i read everything; i answer most of it within a few days.

open to

weird ai use cases, papers worth reading, anything about endangered languages, swiss coffee.