From Prototype to Product - What AI Tools Miss

AI coding tools have made it easier than ever to build a working prototype. But there's a gap between "it works on my screen" and "it's ready for real users" — and that gap is where most projects fall apart. In this session, ex Microsoft software engineer and Second Life Software founder Katia Smith walks through what AI tools consistently miss when generating code, why it matters, and how to recognize where your project actually stands. Drawing from her experience running Second Life Software — where she's taken dozens of vibe-coded prototypes and turned them into production-ready products — Katia breaks down the real journey from prototype to product.

What you'll learn:
The Five Gaps AI Tools Leave Wide Open. Not a tutorial on how to fix these but a tour of what goes wrong and why it matters. Real before-and-after examples from agency work:

• ‍‍Security - AI builds the feature but not the protection. Show real-world consequences: the 380,000 exposed apps, leaked patient records, open databases. What it looks like when nobody's checking who's allowed to access what.
• UX beyond the happy path - AI only builds what the screen looks like when everything works. What does your user see when something fails? When there's no data? When they lose connection? These moments are where trust breaks.
• Secrets in the code - What happens when your API keys are sitting in your source code and bots find them. Real stories of people waking up to thousands of dollars in charges.
• Error handling - What happens when a payment fails? When the API is down? When a user does something unexpected? If nobody planned for it, the app just breaks silently.
• Deployment gaps - The difference between "it works on my laptop" and "it works for 1,000 users at 3am." What changes when your app goes from demo to live.