DecisionPen

The Architecture Context Plane

DecisionPen is the institutional memory of how your enterprise is built and why those choices still matter.

Agents keep that memory current as work happens, and humans review what becomes trusted before it joins the Enterprise Knowledge Graph.

Architecture Decision Workspace. No credit card is required.

Traditional Enterprise Architecture vs DecisionPen

Most IT transformation programmes burn months rediscovering what documents already know. The ambition is the same — know the estate — but DecisionPen uses a different operating model for an agentic world.

  • Getting started

    Traditional EA

    Teams start with manual modelling and workshops, and often wait months before a usable baseline exists.

    DecisionPen

    DecisionPen starts with automated discovery from a CMDB plus unstructured documents such as HLDs.

  • What gets captured

    Traditional EA

    Traditional tools mainly capture thin structural factsheets such as names, owners, and lifecycle status.

    DecisionPen

    DecisionPen captures structure plus evidence — High Level Designs, decisions, requirements, NFRs, security notes, and related institutional memory.

  • Freshness

    Traditional EA

    The repository is often outdated within weeks of being declared complete.

    DecisionPen

    Context stays fresh through continuous publish from where work happens via MCP, while stewards confirm what joins the map.

  • Who updates it

    Traditional EA

    Architects act as data clerks on a quarterly cadence to keep the inventory filled in.

    DecisionPen

    Agents publish updates as delivery evolves, and humans approve what joins the Enterprise Knowledge Graph.

  • EA vs solution design

    Traditional EA

    Enterprise architects and solution architects work in silos: EAs live in the EA tool, while SAs keep designs in SharePoint, Confluence, and similar stores.

    DecisionPen

    Living enterprise architecture and agentic solution architecture share the same Context Plane in Claude, Cursor, and the workspace.

  • Delivery & agents

    Traditional EA

    There is little sync with project delivery, so designs, ADRs, and pipeline work stay outside the EA tool and agents hallucinate when the repository is empty or stale.

    DecisionPen

    The Context Engine (MCP) hooks into the agentic SDLC so design and delivery use current architecture truth.

  • Institutional memory

    Traditional EA

    How and why remain scattered across documents, chat threads, and people’s heads.

    DecisionPen

    How and why stay attached to each entity — including meeting transcripts — and remain citable when you ask.

  • Governance

    Traditional EA

    Success is measured mainly by completeness of the inventory.

    DecisionPen

    Success is measured by curation before Confirmed status: AI proposes, and people decide.

DimensionTraditional EADecisionPen
Getting startedTeams start with manual modelling and workshops, and often wait months before a usable baseline exists.DecisionPen starts with automated discovery from a CMDB plus unstructured documents such as HLDs.
What gets capturedTraditional tools mainly capture thin structural factsheets such as names, owners, and lifecycle status.DecisionPen captures structure plus evidence — High Level Designs, decisions, requirements, NFRs, security notes, and related institutional memory.
FreshnessThe repository is often outdated within weeks of being declared complete.Context stays fresh through continuous publish from where work happens via MCP, while stewards confirm what joins the map.
Who updates itArchitects act as data clerks on a quarterly cadence to keep the inventory filled in.Agents publish updates as delivery evolves, and humans approve what joins the Enterprise Knowledge Graph.
EA vs solution designEnterprise architects and solution architects work in silos: EAs live in the EA tool, while SAs keep designs in SharePoint, Confluence, and similar stores.Living enterprise architecture and agentic solution architecture share the same Context Plane in Claude, Cursor, and the workspace.
Delivery & agentsThere is little sync with project delivery, so designs, ADRs, and pipeline work stay outside the EA tool and agents hallucinate when the repository is empty or stale.The Context Engine (MCP) hooks into the agentic SDLC so design and delivery use current architecture truth.
Institutional memoryHow and why remain scattered across documents, chat threads, and people’s heads.How and why stay attached to each entity — including meeting transcripts — and remain citable when you ask.
GovernanceSuccess is measured mainly by completeness of the inventory.Success is measured by curation before Confirmed status: AI proposes, and people decide.

Read the full argument

One shared Context Plane

Your CMDB, documents, and enterprise architecture model share one Context Plane, and then Claude, Cursor, and SDLC agents connect to it through MCP. Enterprise discovery, institutional memory, agentic SDLC, agentic solution architecture, and living enterprise architecture all run on that same plane.

DecisionPen shared Context PlaneCMDB and unstructured evidence feed a shared Context Plane that fuels enterprise discovery, institutional memory, agentic SDLC, agentic solution architecture, and living enterprise architecture.FEEDS THE PLANECMDB / inventoryStructured identityCapabilities, processes,apps, technology, etc.Unstructured evidenceHLDs, meeting transcripts,URS / NFRs / ADRs,security, agent outputs, etc.CONTEXT PLANEDecisionPenShared Context PlaneContext of RecordEnterprise Knowledge GraphEvidenceDecisions and requirementsEstate changesSAME PLANE FUELSEnterprise discoveryInstitutional memoryAgentic SDLCReqs / NFRs / security / complianceAgentic solution architectureClaude / Cursor / same contextLiving enterprise architectureHumans and agents publish / stewards confirm
CMDB records, documents, and the enterprise architecture model share one Architecture Context Plane, and then Claude, Cursor, and SDLC agents join that same plane. Enterprise discovery, institutional memory, agentic SDLC, agentic solution architecture, and living EA all run here instead of living in three silos that never meet.

Architecture Decision Workspace. No credit card is required.

See the workspace

Watch a short look at DecisionPen as an Architecture Decision Workspace, then try it yourself.

Watch 90 seconds

Think ChatGPT or Gemini — but for your enterprise architecture. See how DecisionPen works as an Architecture Decision Workspace.

Ready to try it yourself?

See it live

Built for anyone who decides under change pressure

  • CTO / CIO

    Move change faster without flying blind across the estate.

  • Enterprise architect

    Maintain a living landscape instead of another failed inventory project.

  • Solution architect

    Design against shared truth in Claude, Cursor, or any other connected agent.

  • Engineering / Operations

    Give delivery agents context they can trust before change ships.

  • Transformation / M&A

    Surface dependencies from institutional memory rather than from blank workshops.

  • Wider organisation

    Get grounded answers and bring people up to speed faster.

Try the Architecture Decision Workspace

See the Context Plane live. Start with the example workspace, then bring a system of interest of your own.

No credit card is required.