Philosophy & genesis

Where TEAF comes from, and why this name

TEAF wasn't born in a research lab. It was born in steering committees, architecture reviews, and diagnostics of industrial organizations and financial firms trying to understand why their technical debt kept growing despite sustained investment. The pattern was always the same: executive committees fund transformation, IT departments deliver projects, and the gap between strategic intent and the reality of systems keeps widening — a gap generative AI accelerates more than it resolves.

The word "Temporal" in the name doesn't refer to scheduling or deadlines. It refers to continuity: an architecture that evolves with time and accumulated experience, rather than one frozen between periodic review cycles. This word choice has no connection to Temporal, the commercial software workflow orchestration platform — two independent uses of the same word.

Central thesis

"An organization stays governable as long as its governance speed matches its execution speed. Below that condition, architecture debt grows faster than the value produced."

This is TEAF's founding principle — a working hypothesis argued from repeated field experience, not a result empirically validated on comparative data. Book I details its foundations and invites the reader to judge it as such.

What TEAF contributes, precisely

Inherited, adapted, combined, unique to TEAF

TEAF doesn't claim eleven isolated terminological inventions. This distinction protects the framework's credibility with a demanding reader.

Inherited

TOGAF, ArchiMate, ITIL, COBIT, SAFe remain the foundations. TEAF doesn't replace them: TOGAF remains the methodological backbone, ArchiMate the reference notation.

Adapted

The Living ADR adapts the Architecture Decision Record practice (popularized by Michael Nygard, 2011) with a permanent link to the Knowledge Backbone. The Digital Twin, already widespread in engineering, is applied here to enterprise architecture.

Combined

Knowledge Graph, continuous governance and AI supervision exist separately elsewhere. No published framework brings them together in a single enterprise architecture system tied to named governance roles.

Unique to TEAF

The closed eight-component Loop — Intent through Learning — and the five named governance roles that run it. No existing framework links these eight steps in a single, closed, end-to-end instrumented cycle.

How the loop works

The eight components of the TEAF cycle

From strategic intent to learning, every cycle closes the loop back to the next intent — without interruption, without re-entry.

01Intent
02Capability
03Knowledge
04Decision
05Automation
06Execution
07Observation
08Learning

The detail

What each component does

The framework extends into eight operational patterns — AI Control Plane, Knowledge Backbone, Decision Fabric, Continuous Governance, Living ADR, Capability Graph, Digital Twin Loop, Architecture Intelligence Layer — detailed across the TEAF collection.

01

Intent

The strategic intent

The board's objective translated into actionable language, traced through to the decision.

02

Capability

The capability engaged

The business, technical or compliance capability engaged to serve an intent.

03

Knowledge

The available knowledge

Data, past decisions, dependencies and constraints — a single source of truth.

04

Decision

The architecture choice

The choice made from available knowledge, formalized as a Living ADR.

05

Automation

The automated execution

The automated execution of a decision, supervised by the AI Control Plane.

06

Execution

The change produced

The change actually produced in the information system.

07

Observation

The measured gap

The measurement of the gap between the outcome achieved and the original intent.

08

Learning

The learning

The learning drawn from observation, fed back into the next intent.

Cross-cutting foundations

Six foundations, constant across every layer

They are not a stage of the program: they apply simultaneously to every component of the loop.

Secure

Zero Trust · IAM · GDPR · DR/BC plan

Govern

Architecture Review Board · Living ADR · Continuous Governance

Data Architecture

Knowledge Backbone · Data Mesh · Vector DB · RAG

Observe

SRE · SLI/SLO · Prometheus · Grafana

Sovereign

Control · Resilience · Hybrid cloud

Legal & Compliance

GDPR · EU AI Act · Audit · Tax

Deployment

Four phases: Why, What, How, Who/When

Each phase produces concrete deliverables before the next one is engaged. No total, simultaneous overhaul of the information system.

Phase 12 to 4 weeks

Why

  • TEAF mission charter
  • Strategic scoping note
  • Prioritized list of formalized Intents
Phase 24 to 8 weeks

What

  • Initial Capability Map
  • Knowledge Graph v0
  • Architecture debt report (IDA)
Phase 32 to 6 months

How

  • Target architecture & AI exoskeleton selection
  • Agent specification & AI Control Plane
  • Allocated compute budget
Phase 4Ongoing

Who/When

  • Architecture Review Board & 5 TEAF roles
  • First active Living ADRs
  • Sequenced deployment calendar

Maturity model

Seven levels, from documentation chaos to the digital nervous system

Honestly positioning the starting point conditions everything else. The online diagnostic places your organization in six minutes.

0

Documentation Chaos

No reliable overview. Every decision starts from zero; knowledge lives in a handful of people's heads.

1

PowerPoint Architecture

Diagrams and doctrine exist on paper but don't drive real decisions.

2

Structured Architecture

Mapping and decisions are documented and kept up to date, but execution remains largely manual.

3

Instrumented Architecture

Decisions are tracked (ADR), deployments are tooled, first indicators are monitored.

4

Connected Architecture

Knowledge, decisions and execution are linked in a single system. Gaps are measured systematically.

5

Living Architecture

The intent → execution → observation → learning loop runs continuously, with governance built in.

6

Digital Nervous System

AI orchestrates most of the cycle under human supervision. The organization learns from every decision in near real time.

Guiding principles

Ten TEAF principles

01

Explicit Intent

Every architecture action starts from a formalized intent, traceable back to the board.

02

Decisional Continuity

An architecture decision is a flow, not a milestone: it is re-evaluated with every relevant new piece of information.

03

Total Traceability

Every technical component links to a documented decision, itself linked to an original intent.

04

Single Source of Knowledge

One living source of truth — the Knowledge Backbone — replaces parallel documentation.

05

Governed Automation

No agent, no pipeline, deploys a critical change without passing through the Decision Fabric.

06

Closed Loop

Observation and learning systematically close the cycle back to the next intent.

07

Continuous Governance

Compliance is assessed continuously, not through point-in-time audits.

08

Adaptive Granularity

The level of detail required adapts to the criticality of the capability, not a uniform standard.

09

Framework Coexistence

TEAF integrates with TOGAF, ITIL or COBIT without replacing them: it orchestrates their continuous execution.

10

Cognitive Sobriety

The complexity shown to a leader always stays below the real complexity of the system.

Explore TEAF through theory

TEAF Full — the complete collection

The four books forming the complete corpus: the Manifesto (thesis and vocabulary, for COMEX and CIOs), the Framework Specification (metamodel, artifacts, roles, for architects), the Practitioner's Guide (deployment, playbooks, for consultants and ops teams) and the Operational Implementation Guide (the 1-1-1 principle, scoping a POC, deployment trajectories). V1 of the collection, now complete — a V2 is planned, enriched with constructive feedback from other organizations, with no release date announced. Available in paperback and Kindle.

Explore TEAF hands-on

TEAF Light — the open source reference implementation

The executable implementation of the 1-1-1 principle detailed in Book IV — one process, one loop, one measurable objective — and a reference stack (Dolibarr, GitLab, PostgreSQL/pgvector, n8n, Python, Cloudflare) to start on a limited scope in a few weeks. Free and open source, unlike the book that documents it; can also serve as a POC before committing to the full collection.

TEAF Light reference architecture diagram: from the TEAF metamodel (loop, governance, metrics) to its implementation via Dolibarr, GitLab and Cloudflare, orchestrated by n8n/Python down to AI agents.
TEAF Light's reference architecture: from the metamodel to its open-source implementation.
TEAF FullTEAF Light
Content4 books: Manifesto, Specification, Practitioner's Guide, Operational Implementation GuideOpen source reference implementation, executable companion to Book IV
GoalUnderstand the full theory, train roles, the Architecture Review BoardStart execution on a limited scope
Principle8-component loop, 10 principles, 9 anti-patterns1-1-1 principle: one process, one loop, one measurable objective
Starting pointReading, strategic framingPOC in a few weeks
AccessAmazon, paperback and KindleGitLab, open source license

Where does your organization stand?

Take the diagnostic or explore the role-based certification program.