Top tier
Editor & metadata
The modern editor and the Dictionary metadata server. This is where engineers and AI agents read authoritative facts about the estate and drive analysis, specification capture and generation.
One canonical view, end to end.
The platform composes as three tiers: editor and metadata at the top, the runtime and integration layer in the middle, and the data layer at the base. The same diagram lives in the documentation and in the demo, so what you are shown is what you would run.
The three tiers
How the platform stacks up.
Middle tier
Runtime & integration
The real COBOL runtime and the .NET integration layer, including the marshalling and calling infrastructure. This is where code actually executes and where legacy meets modern services.
Base tier
Data
The data layer: legacy files and databases, served through the platform’s access providers so the data can modernise independently of the code that reads it.
Metadata, injected at compile time
Custom Attribute Injection (CSATTPP).
Metadata is not bolted on at runtime and hoped for. Compile-time Custom Attribute Injection (CSATTPP) weaves the platform’s metadata into the compiled artefacts themselves, so the structural facts an AI agent relies on travel with the code rather than living in a side file that can drift out of date.
See it in motion
Walk the architecture against your estate.
A working session mapping the three tiers onto the systems you actually have to modernise, and where your estate would sit in each.