Cloud-based feedback capture system for trade shows

To be honest, traveling to Printing United 2025 in Orlando, Florida, I had set my expectations relatively low. But, oh boy, was I wrong. This Printing United reminded me of the best Print shows in Chicago. So many demos, so many qualified customers, and great, great discussions. Typically, I would have my notebook ready to record feedback, interesting feature requests, and those minor issues you only find by doing 50+ demos of your product. But I forgot my notebook, my noise-cancelling headphones, and a few other essential items. While annoying, I was most upset about leaving my notebook on my desk at home. At a typical trade show, you're capturing dozens of customer interactions daily. I was about to downgrade to the hotel's cheap notepad where I knew I'd end up with crumbled pages in my backpack and half-readable notes that make no sense two weeks later. Then I remembered Anthropic had just launched Claude Cloud Code—a cloud-based development environment accessible via the Claude app that lets you create custom workflows on demand. This wasn't just about taking notes anymore. What if I could transform real-time customer feedback directly into structured requirements documents while still manning our booth? The night before the show, I spent 30 minutes setting up a Claude project with our API documentation, node architecture, and design patterns. I created a simple "skill" that would analyze customer requests against our existing codebase. Think of it as giving Claude a crash course in our product before letting it loose on customer feedback. Here's how it worked: A customer would explain their workflow challenge during a demo. After the demo, I'd open Claude on my phone during a brief break and trigger the skill with a quick voice note about their request. Then back to the next demo, the next feature request. Meanwhile, Claude was busy validating the request against our codebase, identifying conflicts, pulling relevant API documentation, and structuring everything into our standard requirements template. One customer wanted integration with a 3rd party cloud based task manager. Previously, I would've scribbled "integration needed - check details later" and hoped I'd remember the context. By evening, Claude had already mapped out the integration points, identified similar requests from other customers, and drafted a complete requirements document suggesting we build a flexible framework rather than one-off solutions. That evening, instead of spending hours reconstructing conversations from messy notes, I spent five minutes reviewing Claude's work, answering its clarification questions, and approving the requirements document. Over 2.5 days at Printing United, this improvised system generated a dozen complete requirements documents—work that usually takes 2-3 weeks post-show. I captured 100% of customer requests versus my typical 60-70%. Most importantly, I discovered connections between similar requests I never would have noticed in the trade show chaos. Remember when I dared you to give AI your PRD in my last article? This is what happens when you flip it—let AI write the PRD directly from customer conversations. The context becomes the product. Your next trade show can be your next customer round table. And with the right AI setup, it can also be your most productive requirements gathering session—even if you forget your notebook.