Building a Business with LLMs: Lessons from the Command Line

Keep your plans in one place. Free account.

Kent Barton discusses building tracingviolet with AI agents, sharing what worked, what did not and how to assess whether generated code works as intended.

AI summary · From the event listing

Anyone Can Build Software Now. So What Will You Build? These days, anyone can build software. So... what will you do with this newfound superpower? AI is changing how software gets built, making it possible for individuals to turn ideas into functioning products without a traditional engineering team. But building with AI comes with its own challenges. How do you know the code actually works when you haven't written it yourself? And what does it really take to go from an idea to a working product? In this session, we'll explore the opportunities, challenges, and practical lessons of building software with AI agents, from moving faster and avoiding common mistakes to knowing when to trust AI and when to question it. SPEAKER Kent Barton | Founder, tracingviolet Kent is the founder of tracingviolet, an agent readiness service that helps publishers of MCP servers and APIs get discovered, chosen, and used by AI. He researched, designed, and built tracingviolet using Claude and Codex as his sole team members. In this talk, Kent will share what worked, what didn't, and what he'd do differently if he were starting over. AI agents are fast, tireless, and confident, but they can also make mind-blowingly stupid mistakes. The real challenge isn't getting code written. It's knowing whether the code actually does what it's supposed to do. You no longer need a team or years of engineering experience to build something real.

Baker