How to break a service down into a factory
A six-step method to turn a service into an AI Factory that delivers margin — before writing a single line of code.
Professional services — consulting, support, application management, BPO — are moving from hourly billing to outcome-based pricing, because AI agents now perform part of the work. This is Service-as-a-Software. Getting there means turning the workshop into a factory, and that takes method.
1 · Diagnosis. Map the current process before designing the new one: who does what, with which tool, where the queues and rework are. Without a baseline — average time, cost per request, recurrence — no later gain can be proven.
2 · Decomposition. Separate intelligence tasks (recognize, classify, retrieve, synthesize) from judgment tasks (decide in context, take accountability, negotiate). AI does the former; people keep the latter. At the boundary, humans act as reviewers.
3 · Applicability. Plot volume against complexity in a 2×2 matrix. The factory delivers most where repetition is high and judgment is low: start with the highest-volume, lowest-complexity quadrant.
4 · Equation. Turn the analysis into a business case — investment, feasibility, impact and time — and calculate the opportunity cost: how much margin is lost every month without the factory.
5 & 6 · Synthesis and validation. Chain everything into blocks, each with a deliverable and a KPI, from diagnosis to a designed pipeline and a factory running in production, validated within 90 days.
“Automate what can be recognized; leave to people what must be judged.”