Docs/How Neuron works

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

Architecture example · Architecture design
  1. Enterprise sources
  2. Source understanding
  3. Schema & concept discovery
  4. Identity resolution
  5. Relationship mapping
  6. Rules & business meaning
  7. Neuron ontology
  8. Context runtime
  9. 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.

Current implementation · Available in this release
  1. File / HTTPS snapshot / MCP discovery
  2. User-selected context item
  3. Saved source version
  4. Source search / invoice checks
  5. Ask / source-update Run
  6. Evidence

Data + knowledge + tools

InputMeaningExample
DataWhat exists and what happenedInvoices and purchase orders
SOPHow work should happenApproval policy
MCPWhat capabilities are availableA 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.