Enterprise architecture that knows what things cost.
A complete EA platform — applications, capabilities, processes, topology — kept current by your pipelines and agents. Because the model is live, every outage, audit and roadmap decision comes back as a number: who is affected, revenue per hour, and which regulation is in scope.
Everything you expect from an EA suite. Graph-native underneath.
Applications, components, dependencies, capabilities, domains, business processes, actors and hosting topology — one model, every view derived from it. Draw a diagram and you have modelled the estate: every node and line is a real asset in the same graph.
Application portfolio
TIME quadrants auto-scored on technical and functional fit. Every score decomposable.
Capabilities & processes
Versioned business processes, linked step by step to the systems they run on.
Hosting topology
Providers, zones and connectivity. Where systems actually run, and how they interconnect.
Tech radar & EOL
End-of-life synced daily and mapped to the components that depend on it.
Then the question no EA suite answers: what does this failure cost?
Click any component. Dependra traverses seven layers, from infrastructure to strategic objectives, and rolls the result up into the language the business speaks.
Clouds, data centres, zones, connectivity
APIs, databases, queues, workers
Systems you build, run or subscribe to
Versioned workflows and their steps
What the organisation delivers
Who is affected, at what cost
Strategic goals, live from the landscape
Your monitoring says the database is down. Dependra says it just put €6,600 an hour of checkout revenue at stake and started the SLA clock on your largest contract.
The model moves at the speed of your code.
Every EA tool dies the same way: the model rots faster than a handful of architects can update it. Dependra removes the chasing. Point a coding agent at a repo over the MCP server, or commit a YAML manifest and let CI publish it.
The paste-in import always previews a diff before it writes. From CI, your pull request is the gate.
Objectives computed from reality, not typed into a dropdown.
Key results are wired to the landscape underneath them. When a component degrades, the signal climbs the graph and lands on the objective it threatens.
Stop rebuilding audit evidence. You already have it.
The model that computes blast radius reads, unchanged, as the artefacts auditors ask for. Unencrypted paths surface as findings. Every claim is backed by a real field.
One prevented outage pays for the year.
Take one revenue-critical capability. Set what an hour of downtime costs, and how many hours a year go to working out who is affected.
The incumbents are excellent. Dependra answers a different question.
We overlap established EA suites heavily, and we say plainly where they are stronger. Here is where we differ.
One model. Start where the pain is sharpest.
Turn every outage and audit into a defensible number.
Load your first landscape and watch a blast radius land as cost per hour, in your currency.