Blog
AI automation, in practice.
Thoughts on AI automation, building in public, and the future of work.
June 12, 2026
n8n, Claude, Gemini, OpenClaw, and Hermes compared (2026): where should a recurring AI task live?
Workflow builders, built-in agent schedulers, and dedicated job layers fit different AI tasks. See how n8n, Claude, Gemini, OpenClaw, Hermes, and Golemry compare, and when an n8n alternative fits.
June 9, 2026
Agent Automations You Don't Have to Babysit
Setting up agent automations is easy now. Trusting them to run unattended is not. The missing piece is a separate overseer that watches each run so you don't have to.
March 20, 2026
Why AI Automations Fail: The Capability-Reliability Gap
AI models keep getting smarter, but not more reliable. The gap between what AI can do and what it does consistently is where most automations break.
March 18, 2026
Why Your AI Agent's Output Gets Worse Over Time
AI automations work on day one, then quietly degrade. The problem isn't the prompt. It's the lack of a feedback loop.
March 17, 2026
From OpenClaw Cron Job to Production Automation: What's Missing
AI cron jobs work in demos but break in production. The missing piece is a reliability layer between execution and delivery.
March 16, 2026
Why I'm Building Golemry
The progression from AI assistants to AI operations, and why the review layer is the missing piece.