Building multi-agent systems as a non-engineer
How I actually design, run, and debug multi-agent AI systems without an engineering background — and what that says about who gets to build now.
I build multi-agent systems. I am not an engineer by training. Those two facts are less in tension than they used to be, and the reason is itself the story: the interface to building has moved up the stack, and judgment about how to decompose a problem now matters more than knowing the syntax to express it.
Here is how I structure the systems I run, where they break, and what I have learned about keeping them reliable enough to depend on.
Draft — placeholder body. Replace with the real write-up. This seed establishes the third featured slot on the home page.