Writing
Papers, experiments, and notes from building with AI.
Why I keep rebuilding my own tools
Notes on building Crossthreads and Headless_CRM, and what shipping your own tooling teaches you that reading about it doesn't.
- SSRN ↗
More Agents, More Problems? Cost, Reliability, and Complexity in Enterprise AI Operations
A cost-aware empirical study across 3,751 recorded-cost result rows and 2,183 reliable operations. Small constrained swarms can improve quality; larger swarms and heavier harnesses do not reliably pay for themselves once cost, reliability, and failure surface are counted.
- Humaie Research ↗
Designing the Intelligent Enterprise
From hierarchy to capability — how agentic AI is obsoleting the org chart, what the new paradigm of human–agent collaboration looks like, and where it leads.