How does EA Workbench fit with existing EA tools?

It complements rather than replaces them. Most organisations keep their system of record for governance and use EA Workbench as the practitioner surface where architecture is actually written and maintained.

Why YAML and git rather than a database?

Because that is where engineering already works. YAML in git makes architecture diff-able, reviewable, and reachable by both humans and AI agents without a custom integration layer.

Insights

Product Sheet · Consid

EA Workbench

AI-powered enterprise architecture. A practitioner-first approach: YAML-based, git-native, and designed to be maintained alongside the code it describes. From weeks to days.

By: Product Sheet — Consid
Published: April 2026

What AI makes possible now

Architecture work at a different pace.

Governance decisions in days, not weeks
Mapping, health assessment, compliance checks, decision preparation — done in days instead of weeks. Repeatable, auditable, consistent.
Architecture as a competitive advantage
Companies that know their landscape make better decisions faster. You get that overview without a six-month programme.
Operationalize your architecture practice
Your architecture processes become codified playbooks. They scale beyond the people who built them.
Audit-ready by default
Your catalog maps systems to regulatory obligations continuously. Compliance becomes a byproduct of how you work, not a separate project.

Three modes

Start with the mode that matches your need.

Discover — What do we have?

Detailed application documentation, a structured catalog connecting systems and data, invisible dependencies made visible, and a health assessment with friction scorecard across six lenses.

Direct — Where are we going?

Strategic architecture direction, principles and standards, a technology radar, gap analysis with prioritized gap register, and a modernisation roadmap based on gap severity and business criticality.

Govern — How do we stay on track?

Structured decision workflow with full traceability, drift detection monitoring, periodic health assessment, and continuous compliance monitoring against regulatory obligations.

Each phase — direction, target architecture, gap analysis — is approved before the next one starts.

Health assessment

Six lenses. Maturity Level 0 to 4.

Knowledge, Structural, Governance, and Guidance look inward at the practice. Compliance maps the architecture against regulatory obligations. Integrity checks whether the connections between business domains and supporting systems are structurally sound.

Knowledge
How complete, current, and well-described the catalog is. Runs first — the reliability of the other five depends on it.
Structural
Coupling and dependencies, legacy concentration, blast radius. Technical debt made visible as architectural risk.
Governance
How decisions are actually being made: whether ownership is clear, authority is defined, and decisions are being documented.
Guidance
Whether principles, standards, and technology decisions exist and are being applied.
Compliance
Coverage against GDPR, NIS2, EU AI Act, ISO 27001. With NIS2 enforceable across the EU and EU AI Act August 2026 deadline approaching, rarely optional.
Integrity
Whether connections between problem and solution space hold up — systems without a domain, data objects without an owner, components without a parent.

Guided by playbooks

Every workflow follows a structured process.

Health assessment
Assess architectural maturity through six friction lenses. Produces the scorecard.
Compliance mapping
Discover relevant obligations, map them to your architecture, identify gaps.
Architecture direction
Define strategic direction, author target guidance, run gap analysis. Phased with approval gates.
Decision governance
Triage, prepare, and record architecture decisions. AI assembles the brief: relevant principles, blast radius, past decisions, compliance implications. Full evidence trail.
Drift detection
Compare current landscape against approved direction. Surface misalignment early.

When we leave, the playbooks stay. They become your organisation's codified operating procedures for architecture work: repeatable, consistent, and independent of who runs them.

EA Workbench vs. alternatives

A living catalog you own.

With EA Workbench	Instead of
Days to first result	Weeks to first result
Living, interactive catalog	Static documents that go stale
A navigable overview you own	A report in a drawer
Codified, reusable playbooks	Ad hoc, person-dependent consulting
Friction scorecard with six lenses	Health assessed ad hoc or not at all
No license or subscription	Expensive EA tooling

Questions

Frequently asked

What is EA Workbench?: An AI-powered enterprise architecture toolkit. Practitioner-first, YAML-based, git-native, and designed to be maintained alongside the code it describes — so architecture stays current rather than ageing in a separate tool.
How does EA Workbench fit with existing EA tools?: It complements rather than replaces them. Most organisations keep their system of record for governance and use EA Workbench as the practitioner surface where architecture is actually written and maintained.
Why YAML and git rather than a database?: Because that is where engineering already works. YAML in git makes architecture diff-able, reviewable, and reachable by both humans and AI agents without a custom integration layer.

3 applications in a proof sprint.

Pick three applications — systems that matter, ideally where documentation is incomplete or dependencies are unclear. Within one week: structured catalog entries, a dependency map, an initial health observation, and a concrete plan for what a full engagement would cover.

Talk to Consid