We started Nextx Ventures after eight years operating Üretken Akademi — Türkiye's largest AI-powered virtual internship platform.
What we learned in those years: most studios produce pretty pitch decks, most agencies bill hours and ship features, but very few actually operate what they build at scale. We wanted to be operators — to run our products to real users, see them through, and bring that same discipline to founders building the next AI wave.
So we did. Üretken Akademi crossed 85,000 users with corporate simulation partners like Turkish Airlines, NVIDIA, İş Bankası, and Enerjisa. UNDP added the founder to their Global Policy Network ExpRes Roster. We ran 90+ programs across Türkiye, Azerbaijan, Dubai, and Egypt with 100+ institutional partners. Models have been in our production loop from day one.
Nextx is the studio version of that operating discipline — extended outward. We co-build with founders, embed as fractional CTOs, and ship AI the way it actually scales: eval-first, production-grade, operationally serious.
EdTech operator since 17. Founded Scode (mobile coding platform for kids — 450k+ downloads, Google's Best Apps of 2017) before co-founding Üretken Akademi, Türkiye's flagship AI-powered virtual internship platform with 85k+ users and 100+ institutional partners. Approved consultant on UNDP's Global Policy Network ExpRes Roster, advising digital innovation and entrepreneurship across global development projects. Y Combinator Startup School alum.
LinkedIn →Approved expert on the ExpRes Roster — advising Digital Innovation & Entrepreneurship across global development projects aligned with the SDGs.
Awarded to Scode, the founder's first mobile platform — a children's coding app that reached 450k+ downloads and a 4.5/5 store rating.
Founder is a Y Combinator Startup School alum from the Scode era — same operating playbook informs everything we ship today.
Turkish Airlines · NVIDIA · İş Bankası · Enerjisa · Albaraka · UNDP · Google Developers · Teknopark İstanbul · Bilişim Vadisi — across Türkiye, Azerbaijan, Dubai, and Egypt.
We start with the business outcome, not the tech. What does success look like in 6 months? What can't be true at the end?
We sketch the product, the model strategy, and the eval surface together. We don't build until we agree on what 'working' means.
Tight, eval-driven sprints. Every release goes through real production traffic on day one — no staging-only victory laps.
We don't hand off and disappear. We run what we ship, instrument it, and iterate based on what users actually do — not what they say.
Real users on real traffic before any victory laps.
They get the same operational discipline as your database — versioned, monitored, and rollback-able.
If we can't measure it, we won't claim it.
Every choice is informed by what it costs to run, not just what it costs to build.