A new hybrid role called the AI Operator automates business processes using LLM APIs, MCPs, and agent frameworks, rotating across functions every quarter.
Key Takeaways
The AI Operator requires three skill layers: Python and LLM API proficiency, functional business understanding, and high-EQ low-ego execution.
Work cadence is two-week sprints: two days learning a function, one day scoping with the CEO, five days building, two days handing off.
Target automation candidates are deliberately mundane: inbound lead response, post-call Salesforce entries, revenue aging reports, legal review via GC.ai or Harvey.
The steam engine to electricity analogy frames agents and MCPs as requiring full org redesign, not just tool substitution.
Key metric is dollar revenue per employee, tracked alongside AI usage per employee and tasks fully automated.
