Architecture overview
How enterprise information becomes usable business context.
The context layer
Connecting a system gives software access to information. PatternLab’s architecture adds business concepts, shared identity, meaningful relationships and rules between the source systems and Ask or Run. Original enterprise systems remain the sources of record.
The design below explains the intended context engine. Current capabilities are shown separately so the diagram is not mistaken for an implemented automatic pipeline.
Architecture design
- Enterprise sources
- Source understanding
- Schema & concept discovery
- Identity resolution
- Relationship mapping
- Rules & business meaning
- Neuron ontology
- Context runtime
- Ask / Run
Available in this release
Workspaces persist source snapshots, instructions, results and evidence. The canvas uses task templates. Automatic concept mapping, cross-system identity resolution, ontology traversal and AI interpretation are not implemented.
- File / HTTPS snapshot / MCP discovery
- User-selected context item
- Saved source version
- Source search / invoice checks
- Ask / source-update Run
- Evidence
Data + knowledge + tools
| Input | Meaning | Example |
|---|---|---|
| Data | What exists and what happened | Invoices and purchase orders |
| SOP | How work should happen | Approval policy |
| MCP | What capabilities are available | A server describing search or action tools |
Understand before integrating
You can use the current source workflows without learning the ontology design. Read the technical concepts when you need to evaluate how persistent business meaning could support more complex reasoning.