Build · Assess & Decide · Analysis & Finance

AIaaS Productization Decision Register

A founder deciding which of their internal AI models to package and sell as AIaaS gets a defensible, rule-derived register that says Go / Go-with-notes / Stop per model and names the single best candidate to commercialize.

You receive: A JSON object {ideaName, models:[{name, input, output, dataVariation:int 1-3, logicVariation:int 1-3, interCustomerVariability:enum[low,medium,high], processMaturityBand:enum[low,medium,high], systemMaturityBand:enum[low,medium,high], technicalFeasibility:enum[notFeasible,lowFeasibility,mediumFeasibility,highFeasibility], businessPotentialScore:int 1-4, competitivenessScore:int 1-4, productizeVerdict:enum[go,goWithNotes,stop], rationale:string}], recommendedModel:string}

Part of Choose Business Model

What's verified: STUD verifies that the register is structurally complete, every score is in range, and every derived verdict (variability, feasibility, productize) correctly follows the book's lookup tables from the buyer-supplied inputs. STUD does NOT verify that the maturity/variability/business-potential inputs are factually accurate for the real models, nor that the recommended model will actually succeed commercially.