About

I'm Dakota Secula-Rosell. I build infrastructure around AI agents: memory, tools, evaluation environments, deployment loops, and the operational glue that makes autonomous systems useful in production.

Dispatch is where I write field notes from that work — what shipped, what broke, what I misunderstood, and what changed after putting real agents against real systems. The throughline is proof of work over takes.

The specific tools change. The questions stay pretty stable: how should agents use software, how should software expose itself to agents, and what reliability habits matter once agents have real access?

If you want to get in touch, send me an email. You can also find me on GitHub.