About CPT

What this is

A diagnostic engine, not a chatbot.

Built on Constraint Productisation Theory v1.3 — Donald Oluwasegun Adeniji, Institute of Productisation, Lagos, Nigeria.

The CPT formal core
  • Validity Condition — C ⟺ (U ≥ θ) ∧ R. A constraint is valid if and only if it meets the utility threshold and is realisable.

  • Invariance Condition — Reproducibility(U) ≥ V. A constraint is productised when its utility reproduces across environmental variability.

  • Stability Ratio — SR = (CIₛ + CIᵢ) / AF. SR ≥ 1 means constraint integrity exceeds amplification — the system is stable.

  • Fragmentation Threshold — Σ|I| > κ. When constraint interactions exceed structural capacity, the unified architecture fragments.

  • Entropy of Constraint — H(C) = lim (1/T) ∫|dU/dt| dt. The long-run rate of change of a constraint's utility.

Canon

Canonical sources: CPT v1.3 — Dynamic Extensions and CPT Notation Reference(Adeniji, 2026). Failure data preserves the original 8-failure taxonomy verbatim from the prototype authored by the Institute of Productisation.

The premise

Most "failure" frameworks describe symptoms. CPT names the underlying constraint pathology — the structural reason a system stops producing reliable outputs. Once you know which constraint failed and why, the redesign pattern is determined, not invented.

How the engine decides

The diagnosis is deterministic. Selected symptoms are scored against the canonical symptom map; each failure has a known set of canonical symptoms (weight 1.0) and a small number of cross-failure overlaps (weight 0.25). The same input always produces the same output — a property no LLM can guarantee.

The three layers

This product is built as a three-layer system:

  • Knowledge graph — failures, symptoms, causes, patterns, and CPT primitives as nodes; indicates, caused_by, fixed_by, affects as edges. Today this lives as JSON inside the app.

  • Deterministic engine — symptom scoring, validity checks, SR calculation. The brain.

  • Language layer (future) — an optional LLM interface that translates free-text descriptions into structured symptoms and writes natural-language reports. Never the decision-maker.

Privacy

Nothing is sent to a server. Your last diagnosis is stored only in your browser. Clear it any time by clearing site data.

Our team

Our strength lies in our individuality. Set up by Esther Bryce, the team strives to bring in the best talent in various fields, from architecture to interior design and sales.

woman wearing black scoop-neck long-sleeved shirt
woman wearing black scoop-neck long-sleeved shirt
Esther Bryce

Founder / Interior designer

woman in black blazer with brown hair
woman in black blazer with brown hair
Lianne Wilson

Broker

man standing near white wall
man standing near white wall
Jaden Smith

Architect

woman smiling wearing denim jacket
woman smiling wearing denim jacket
Jessica Kim

Photographer