Context Modeler
Context Modeler
Build and explore the knowledge graph that powers your AI agent.
The Context Modeler is how you tell Digel about your plant. It has two surfaces:
Process Canvas
A visual editor for drawing your asset hierarchy and process flows. This is where you build and edit the model.
Knowledge Graph Explorer
A force-directed graph view of everything in the knowledge graph: assets, documents, issues, sensor tags, and all their connections.
What the context graph contains
The graph stores several types of nodes and their relationships:
- Assets, meaning any physical or logical part of your plant: a factory, a production line, a section, a machine, or a custom type you define
- Tags, the sensor data points attached to assets (temperature, flow rate, vibration)
- Documents like PDFs, manuals, and SOPs, indexed for semantic search
- Issues and work orders, linked to the assets they concern
- Notes for free-text observations and tribal knowledge
The AI agent traverses this graph when answering questions. An asset's connected documents, sensor history, open issues, and notes are all within reach of a single conversation.