Yesterday, Text2Test took the stage at Webrazzi AI 2026 in Istanbul. Our Co-Founder and CEO Esin Yilmaz and Chief Software Architect Ekin Gun Oncu presented in front of one of Turkey's most engaged communities of founders, investors, and technology leaders.
The question they brought to the room was the one every engineering team is quietly grappling with right now: AI is writing more and more of the code. So who is verifying that it actually works?
The problem no one is talking about loudly enough
Despite massive investment in DevOps and AI tooling, software testing has become the number one bottleneck to shipping at the speed AI now makes possible. Enterprises are shipping weekly or daily. Manual regression testing cannot keep pace. Testing costs have ballooned to roughly 50% of development budgets. Maintenance overhead runs 40 to 60%. QA engineers are scarce and expensive, and most business teams have no way to contribute to test authorship at all.
The result: slower time to market, high spend with low ROI, and a testing function that does not scale with the speed of development.
This is the bottleneck Text2Test was built to solve.
What Text2Test does
Text2Test is an AI-native test automation platform. It takes natural language test scenarios from any stakeholder, not just engineers, and turns them into intelligently self-healing automated tests that run across web and mobile.
The core flow is simple: Write in plain language. Text2Test compiles using GenAI-powered semantic interpretation of test scenarios and screens. Then it runs, with tests that adapt automatically when the product changes.
The result is a 10x faster development cycle, minimal operational overhead, and a testing process that no longer depends exclusively on QA engineers to author every test.
The team behind it
The credibility of the platform comes from the depth of the team. Esin Yilmaz brings 27 years of experience in technology, data, and AI, with a background as a former CAO and experience at Microsoft, Oracle, and EY. Prof. Cemal Yilmaz, Co-Founder and CTO, is a professor at Sabanci University, holds a PhD from the University of Maryland, and spent years as an IBM researcher with more than 7 years specifically in software testing research and development.
Combined, the founding team brings over 80 years of experience in technology and more than 50 years specifically in AI and research. This is not a pivot into QA. It is what the team has been building toward for decades.
Traction that speaks for itself
Text2Test is not a product looking for its first customer. The platform is already in active deployment at the number one private bank in Turkey, serving 25 million customers, and at a major multinational insurer with 5 million customers. Usage has expanded 7x in six months. There are 10 active proof of concepts underway at companies collectively serving 65 million customers, with 4 in procurement stage.
The impact has been consistent: 10x faster development cycles, validated across major banks. Maintenance that used to require dedicated QA headcount no longer dominates the roadmap.
Why this moment matters
The Webrazzi AI 2026 stage was the right room for this conversation. As AI accelerates code generation across the industry, the gap between how fast teams can build and how confidently they can ship is widening. Testing is the missing layer. Not testing as a checkbox at the end of a sprint, but testing wired into the development process from the moment a requirement is written.
That is exactly what Text2Test provides. And as the AI era matures, it is exactly what every team building serious software will need.
We are grateful to Webrazzi for the platform and to everyone in the room for the quality of the conversations that followed.