Skip to main content

Flows: The Heart of LearningFlow

What Is a Flow?​

A Flow is the fundamental building block of LearningFlow.ai. It's a drag-and-drop, visual, node-based sequence that defines everything about your learning activity, lesson, app, or assessment. With flows, you can build dynamic, interactive, and AI-orchestrated learning experiences without code.

After you enter a Collection and select or create a flow, you launch into the Flow Editor:

  • Node Toolbox (Left Panel): Browse, search, or right-click anywhere on the canvas to add the vast selection of nodes (text, images, interaction, AI, logic, more).
  • Canvas: Build your flow visually. Drag nodes on, connect them, zoom with the +/- buttons or mouse wheel, and pan by dragging the background.
  • Color-coded Connections:
    • Red: Execution (controls what step runs next)
    • Green: Text data
    • Pink: Image data
    • Blue: Audio data
    • Yellow: Knowledge data (RAG)
    • Light blue: JSON data
    • Connections only work between matching colors (enforces data logic).

Building Your Flow​

1. Start with the Start Node​

  • Every flow begins with a Start node, defining where the execution starts.

2. Add and Connect Nodes​

  • Drag nodes onto the canvas or use right-click "search and add" for speed.
  • Connect nodes to define the execution and data flow. Execution (red) connects directionally left to right defining how learners advance.

3. Branching and Merging​

  • Some nodes, like Buttons or Scenario Branch, allow you to split the execution into multiple pathways (e.g., five button choices = five possible routes).
  • Use the Merger node to rejoin split paths back into a single route.

4. Node Types​

  • Content & Display: Show rich text, images, audio, or HTML.
  • Input: Collect text/audio/image, quiz/widget responses.
  • AI Orchestration: Logic using LLMs or Vision models (Knowledge Bot, Multi-Turn Conversation, Scenario Branch, Output Splitter, etc.).
  • Branch/Merge: Shape learner path (branching, merging, checkpointing).
  • Utility: Progress checkpoints, random number generators, JSON/data utilities, and more.
  • Annotation: Command/Comment nodes for annotation and organization customizable color for clarity.

5. Moving and Managing​

  • Easily move, rename, or delete any node.
  • Annotate the flow with command blocks for clarity.

Authoring Features​

  • Instant Zoom/Pan: Mouse controls or onscreen zoom controls.
  • Right-Click Add/Search: No need to drag everything across a giant canvas.
  • Live Preview: Click Run, your flow slides in from the right for full interaction testing and review. Close to return to editor.
  • Version Control: Every save creates a new, restorable version. Roll back at any time.
  • Export/Import: Export your flow as a .flow file for backup, sharing, or reuse. Importing restores a flow and will overwrite the current one.
  • Collaboration: Flows respect collection collaboration rights for team authoring.

AI Orchestration​

  • Certain nodes let you select from multiple AI models (OpenAI, Anthropic Claude, Gemini, etc.).
  • Each AI-driven node offers model selection and the ability to fine-tune the prompt, instructions, and objectives.
  • Costs for AI usage are model-specific (see Credit Rates).
  • Flows can use AI results to drive scenario branches, adapt next steps, or generate/report outputs for teachers and learners.

Execution Path and Data Logic​

  • Execution Flow: Runs left-to-right, step by step, unless split by branching nodes.
  • Conditional Logic: Execution can follow user input, AI output, or random events.
  • Merging: Use merge nodes to re-join divergence from branches back to a streamlined path.

Example: Quiz Flow with AI Summarization & Branching​

  1. Start Node
  2. Display: Shows instructions using Markdown; visually styled as you want
  3. Quiz Widget: Presents either an AI-generated or hand-authored quiz
  4. AI General Prompt: Summarizes user answers for teacher review
  5. XAPI Reporting Node: Saves attempts/results to your LRS
  6. Branch/Continue: Feedback & next steps, depending on AI output or learner answers

Best Practices​

  • Keep flows organized use comments and color optics for clarity, especially as complexity grows.
  • Test regularly use the run/preview function to find logic holes early.
  • Branch meaningfully personalize the learning journey without confusing paths.
  • Leverage AI use orchestration nodes to automate, adapt, and personalize but always review critical outputs.

For more, see: