What I Learned from Hosting Aytekin Tank (CEO Jotform) in AI Meetup Ankara
For this session, our guest is Aytekin Tank, the founder and CEO of Jotform and one of the most respected names in product-led growth (PLG).
I really enjoyed being part of AI Meetup Ankara and having the chance to connect with so many people who are passionate about AI and product building.
You can watch the full event recording here:
For this session, our guest is Aytekin Tank, the founder and CEO of Jotform and one of the most respected names in product-led growth (PLG). With decades of experience building and scaling Jotform into a global product, he brings a unique, hands-on perspective on how to build great products and now, how to build great AI-powered experiences.
Session #2 – Keynote: How to Build a Good AI Product
Speaker: Aytekin Tank – Founder & CEO, Jotform
Aytekin Tank is the founder and CEO of Jotform, a globally used online form and automation platform with tens of millions of users and hundreds of team members across multiple countries. Over the last two decades, he has focused on one core question: “How do you build a genuinely good product?”—and more recently, “How do you build a good AI product?”
In his keynote, Aytekin explained why AI products play by completely different rules than traditional software:
- AI is non-deterministic (it doesn’t always produce the same result),
- it makes mistakes and hallucinates,
- and yet user expectations are extremely high.
He shared how Jotform increased the resolution rate of its AI support agent from around 25% to 75% by:
- reviewing thousands of real customer–AI conversations every day,
- clustering failure cases and systematically improving RAG, documentation, and answer quality,
- and letting the AI call internal tools and APIs so it can actually solve problems, not just “chat”.
Aytekin’s core message: treat AI like a junior teammate—review its work continually, correct it with real data, and iterate fast instead of waiting for a “perfect” first version.
Why You Chose the Jotform Brand for Our AI Products? I Asked Aytekin Tank, and He Answered with Data
In our AI Meetup Ankara session, I asked Aytekin Tank a strategic question I’d been genuinely curious about: “Why did you launch Jotform AI Agent under the Jotform name instead of creating a separate brand?” His answer was a great reminder of how much real user data should guide branding decisions.
Aytekin Tank explained that they actually tested a standalone version first, but it reached only about 1,000 active users. Once they positioned the product inside the Jotform ecosystem, adoption accelerated to roughly 17,000 active users. The reason was simple but powerful: Jotform’s existing users—many of whom already own websites—were the perfect audience for adding an AI agent directly to their sites.
Session #1 – AI-Powered Customer Support & Sales with LiveChatAI
Speaker: Emre Elbeyoglu – Co-founder, LiveChatAI & Popupsmart
I’m Emre Elbeyoglu, co-founder of LiveChatAI and Popupsmart, and part of the team behind Flatart, a digital marketing agency that has been active for over a decade. We organize AI meetups in Ankara to bring together developers, marketers, founders and technology enthusiasts around practical AI use cases.
In my talk, I shared how our products evolved from an agency need into two B2B SaaS solutions:
- Popupsmart – a conversion optimization platform offering on-site popups and gamified lead capture,
- LiveChatAI.com – an AI chatbot platform that trains on a company’s website, documents and customer data, then answers questions on web, WhatsApp, Shopify and other channels.
I walked through real e-commerce scenarios (inspired by Amazon Rufus) showing how an AI assistant can:
- understand very specific, natural language requests and turn them into product recommendations,
- provide personalized experiences using customer data (orders, birthdays, total spend),
- trigger AI actions via APIs (order status, CRM lookups, dynamic operations),
- and significantly reduce support load while increasing customer satisfaction.
I also touched on how we approach privacy, security and data isolation, how we improve answer quality over time with feedback and better documentation, and ended with a live WhatsApp demo where the audience’s messages were processed by our AI and visualized in real time.