TensorifyDocs

Core Concepts

These are the core concepts that power Tensorify.

Workflows

A workflow is a visual graph that models how your automation logic runs.

A workflow is a visual graph that models your automation logic. In Tensorify, workflows are:

  • edited in the app
  • executed locally
  • tested at selected nodes
  • inspected through outputs and variables

Nodes

Nodes represent steps in the flow.

Examples:

  • webhook trigger
  • HTTP request
  • transform step
  • conditional branch
  • notification or API update

Edges

Edges connect nodes and define how execution and data move through the graph.

Edges define how execution and data move through the graph, ensuring your flow reaches the expected destinations with the correct data.

Selected-node testing

Selected-node testing

The test loop allows you to:

  1. select a node
  2. trigger a test from the app
  3. run locally up to that node
  4. inspect the result

Local execution

Tensorify is local-first for execution.

The product loop centers on commands like:

tensorify watch <workflowId>

Other commands such as run, export, and clone are also part of the current local model.

Variables and outputs

The debugger-style experience depends on being able to inspect:

  • node outputs
  • workflow variables
  • stopped-run state

This provides high visibility into the internal state of your automation.

Next