How Senior Engineers Build AI Agents That Actually Work

Try Granola for free: https://www.granola.ai/?via=madeline-zhang AI agents can already write code, use tools, browse files, and run commands. So why do they still struggle with real engineering tasks? The problem isn't just writing code. It's making AI agents reliable, letting them work for hours without constant supervision, and coordinating multiple agents on large, complex projects. In this video, I break down the three key concepts behind modern AI agent engineering: harness engineering, loop engineering, and graph engineering. You'll learn how they work, how they fit together, and when to