thinking
Writing I come back to. On AI architecture, product, and building things that actually work.
A lot of companies focus on optimizing the wrong part of their AI stack. I know because I made the same mistake early on. Here's why the orchestration layer is where the real value lives.
ai orchestration architecture
Two years building with Claude Code taught me what actually changed. The developer role split into Architect-Operator and AI Workflow Engineer. Here's the framework for staying valuable.
developer role 2026
The €300B enterprise SaaS market was designed for people using dashboards. AI agents use endpoints instead. This shift has direct implications for HR platforms and per-seat pricing.
when ai agents become the buyer
My dotclaude repo is live on GitHub: hooks, agents, skills.
dotclaude