Execution Is Cheap. What now? How to think of AI native orgs | Csaba Tamas | Productlab Conf 2026

📺 PRODUCTLAB CONFERENCE, Berlin 2026 Productlab Conference 2026 brought around 1K+ product people to Berlin and online for three days: Leaders Lab for senior product leaders on 15 Sept, hands-on AI-Craft Day workshops on 16 Sept, and Conference Day on 17 Sept across three stages at The Social Hub and CIC Berlin. The thread running through the week: AI made individuals faster, but organisations, judgment and context are where the real work is now. ✨ About the talk Parloa now generates around 97% of its code with AI, and Csaba Tamas, its CPTO, explains why that produced only a marginal increase in overall delivery speed: coding is one stage, while definition, design, review, testing and alignment didn't get faster. AI brought down the cost of execution, not the cost of coordination. Using the story of Kelly Johnson and Lockheed's Skunk Works, he lays out what an AI-native company should change: cut decision latency, keep requirements rigid while implementation stays scrappy, organise teams around software context boundaries and measure how long decisions take. He closes with a spec- and prototype-first workflow and five metrics for judging the return on AI spend, from idea-to-production lead time to change failure rate. 🎤 About the Speaker Csaba Tamas is CPTO at Parloa, leading product management, UX and enablement, and sits on the Productlab Conference 2026 board. Before Parloa he held senior roles at AWS, advising top fintechs on AI/ML strategy and leading global go-to-market strategy. He was also CPO/CTO at FinLeap Connect and held leadership roles at KEBA, scaling digital products across banking and mobility. He brings 15+ years of experience, an MBA and data science studies at MIT. 🌐 More talks, slides and Productlab Conference 2027 (8-10 Sept, Berlin): https://productlab.app 🎥 Video: Aleksandra Kononchenko