Why AI Apps Fail + Building High-Precision Durable Timers + Confidence Is a Lie

📌 AI Apps Don’t Fail Because of the Tech Alone - They Fail When People Don’t Adopt Them with Andjelka Djukic A great demo gets attention. Getting people to use your AI product in their everyday work is a different challenge. Drawing on practical examples from SaaS, EdTech, and digital marketing, Andjelka Djukic explores where adoption breaks down: confusing onboarding, poor workflow fit, lack of trust, and the gaps between what teams build, sell, and users actually need. We’ll look at how product and development teams can spot these problems earlier and build adoption into delivery, with time for discussion and experiences from the room. Andjelka Djukic is Head of Sales and Marketing at Eclincher and Co-Founder of Innovative Women of Serbia. Her work across SaaS, education technology, and international business development gives her a close-up view of what happens when technology meets real working habits. With a background in education, she brings a perspective on learning, trust, and organizational culture to conversations about AI adoption. 📌 Polling Is a Backstop, Not a Clock: Building High-Precision Durable Timers with Oluwafemi Shobande Backend systems are full of timers: retries, TTLs, leases, scheduled jobs, sleeps. A common way to find what is due is to run a loop that wakes up on a fixed interval and checks the database. That ties precision and cost together. Poll more often and timers fire closer to their deadline, but every replica does more database work, even when nothing is happening. This talk is about breaking that link. The ideas are inspired by firm timers, first proposed for Linux in 2002, and by applying similar ideas at the application layer in Aiki, an open source durable workflow framework created by me. But the design applies to any service that needs to fire things on time. Oluwafemi is the creator of Aiki, an open-source durable execution framework. Before working on Aiki full-time, he was a senior engineer at Bolt, where he…