Improving AI Outputs with Error Analysis 🛠️

AI outputs often look good at first glance, but when we read them carefully, we find mistakes, missing context, weak reasoning, or inconsistent quality. In this session, we’ll learn a practical way to improve AI outputs using human-led error analysis. We’ll discuss: I’ll share examples from my own projects, including converting long technical talks into blogs, chatbots, and other AI workflows. This is a beginner-friendly practical, hands-on session for people building with AI who want more reliable outputs. You need to be familiar with one of these tools: Claude Code, Codex or Antigravity to create your own workflow during the session. Resources: What to bring Note: No charge. Please order food or a drink to support the venue. Organised by AI Study Group: 1 yr | 45+ Meetups | 950+ Members