Nodal’s AI operating model
Updated October 5, 2026
Nodal is a proprietary AI operating model that brings workspace context, agent coordination, inference, and permissioned tools into one system for getting work done.
How Nodal operates
- Context. A request is grounded in the workspace records and connected sources the user and agent are allowed to access.
- Coordination. Nodal combines agent roles, instructions, and task state to organize the work and retain continuity.
- Execution. The runtime uses available inference and authorized tools, subject to connection, permission, and approval requirements.
- Results. Tool results, task status, and recorded usage make completed work and unresolved steps visible.
What is proprietary
Nodal develops the software that connects context, agent roles, work state, inference, and tools into a shared operating environment. The proprietary layer is how those parts work together across a workspace.
The system and its underlying models
Operating model describes Nodal’s software architecture and execution workflow. Underlying language models retain their own ownership and licenses; this is not a claim that Nodal trained a new foundation model or owns third-party model weights.
Nodal supports a private inference runtime and explicit external-model choices. Availability depends on the deployed runtime, connected services, and workspace configuration. A configured model or connection alone does not establish that a task has executed successfully.
Your data and control
Workspace context is used to carry out authorized work. Retaining task context is distinct from training model weights. Nodal does not train foundation models on customer workspace content. Actions remain subject to applicable permissions and approval gates.