Founder
Matin Mavaddat
Epistemic Architect
Over more than two decades in software engineering, security, and systems architecture, Matin has designed and scaled complex systems, including as a Principal Engineer at Amazon. He is the sole inventor on two granted patents, with additional patent applications, and the author of Constructing Systemic Integrity: Epistemic Engineering for the Age of Agentic AI (working title), forthcoming from Oxford University Press.
His Epistemic Engineering framework addresses epistemic debt: the accumulation of unsupported assumptions, unresolved contradictions, ambiguous intent, disconnected evidence, divergent meanings, and lost rationale. He founded Architourge to address the broader problem that increasingly capable people and AI can intervene in complex systems without a coherent understanding of the whole. Artificial Apperception operationalises that work by using model intelligence to develop systemic comprehension and make the evolving systemic representation explicit in the Appercept.
The intellectual journey
From recurring failures to a framework for systemic understanding.
Knowledge is material for investigation, not a synonym for comprehension. Systemic understanding must be constructed before a quality architecture can be deliberately designed and assured.
- 01
Seeing the pattern
Across complex systems, Matin saw consequential decisions outlive the evidence behind them. Teams could explain what a system did, but not always why it had become that way.
- 02
Naming epistemic debt
He came to describe the accumulation of unsupported assumptions, unresolved contradictions, ambiguous intent, disconnected evidence, divergent meanings, and lost rationale as epistemic debt.
- 03
Developing Epistemic Engineering
Matin developed the approach to Epistemic Engineering presented in his forthcoming book: a framework for constructing the systemic understanding required to design and evolve consequential systems. It investigates evidence and competing perspectives, makes epistemic debt explicit and addressable, preserves uncertainty, and clarifies intended and unacceptable consequences before unresolved debt can harden into architecture and code.
- 04
Building Architourge
Matin founded Architourge to address the problem that increasingly capable people and AI can intervene in complex systems without maintaining a coherent understanding of the whole. Its Artificial Apperception Engine operationalises Epistemic Engineering by using model intelligence to investigate distributed knowledge, test perspectives, evolve the Appercept, and develop systemic comprehension that can guide architecture, assurance, change, and implementation.
The work behind Architourge
Practitioner. Inventor. Author. Founder.
His work brings together more than two decades of practice in complex systems, original invention, academic study, and product development.
- 01Practitioner
- More than two decades designing large-scale software systems, including as a Principal Engineer at Amazon.
- 02Inventor
- Sole inventor on two granted patents, with additional patent applications.
- 03Author
- Constructing Systemic Integrity: Epistemic Engineering for the Age of Agentic AI (working title), forthcoming from Oxford University Press.
- 04Founder
- Founder of Architourge, developing Artificial Apperception as the process through which AI can achieve systemic comprehension of complex systems before consequential action.
Academic foundations
Computer science, software engineering, and systems.
- University of Oxford
- MSc in Systems and Software Engineering
- UWE Bristol
- PhD in Computer Science · MSc in Software Engineering
- Bahá’í Institute for Higher Education (BIHE)
- BSc in Computer Science
- Stanford University
- Advanced Computer Security certificate
Why now
As code becomes easier to produce, systemic understanding matters more.
Faster implementation does not eliminate epistemic debt; it can embed that debt more quickly. Artificial Apperception uses model intelligence to discover what must be understood, test evidence and perspectives, and continually evolve the Appercept so that systemic comprehension can inform architecture, assurance, and implementation without losing the reasoning on which the system depends.