Getting the most out of AI for Application Reliability

AI keeps promising to “write 90% of your code” and “10x your productivity.” That sounds exciting… and a bit terrifying. In this live-coding session, Stefan Judis shows what this actually looks like when you care about reliability: using AI agents to spin up real-world monitoring for a Next.js webshop app, without giving up control.

You’ll see how he goes from a plain codebase to a full monitoring setup with uptime checks, API checks, and Playwright Check Suites, mostly by talking to an LLM.

What you’ll see in the recording

• Turn “AI hype” into actual monitors
• Set up uptime monitoring in a few prompts
• Reuse existing Playwright tests as production monitors
• Generate new Playwright tests with MCP
• Debug failures with AI assistance