How to use the T4R Learning Journey Design Agent

The tool walks you through seven steps to design a Local Digital Twin training journey from 72 Micro-Learning Units (MLUs). There are three ways to fill in the first steps — pick the one that suits you. In every path, the AI assists and you decide.

Path 1

On your own

You fill in steps 1–4 yourself, with hints and a glossary to help. In steps 5 and 7 the tool’s built-in AI (Mistral) makes suggestions. You review, edit and confirm everything.

  1. 1Organisational GoalsYou
  2. 2Learner ProfileYou
  3. 3Future RolesYou
  4. 4Current KnowledgeYou
  5. 5Suggested MLUsTool’s AI suggests
  6. 6Finalise MLUsYou decide
  7. 7Assemble JourneyTool’s AI drafts → you decide

Best if you want to think it through step by step, or don’t use an AI assistant.

Path 2 Recommended if you already use an assistant

With your AI assistant (MCP)

Connect your own AI assistant. It asks you questions and pre-fills steps 1–3. For step 4 it sends a summary of what the learner already knows (for example from a CV), and the tool marks the matching MLUs. You review and confirm each step here in the tool. Steps 5 and 7, and every confirmation, happen only in the tool.

  1. 1Organisational GoalsAssistant drafts → you confirm
  2. 2Learner ProfileAssistant drafts → you confirm
  3. 3Future RolesAssistant drafts → you confirm
  4. 4Current KnowledgeAssistant sends a summary → tool marks MLUs → you confirm
  5. 5Suggested MLUsTool’s AI suggests
  6. 6Finalise MLUsYou decide
  7. 7Assemble JourneyTool’s AI drafts → you decide

Supported assistants:

  • Mistral Vibe
  • Claude
  • ChatGPT
  • Other MCP-compatible apps

Server URL: https://learning-journey.k8.design/mcp

Path 3

With any chatbot, no connector

Copy a prepared prompt from the tool into any chatbot and answer its questions. Paste the chatbot’s JSON answer into the tool’s import box, then review and confirm as usual.

  1. 1Organisational GoalsChatbot drafts → you confirm
  2. 2Learner ProfileChatbot drafts → you confirm
  3. 3Future RolesChatbot drafts → you confirm
  4. 4Current KnowledgeYou
  5. 5Suggested MLUsTool’s AI suggests
  6. 6Finalise MLUsYou decide
  7. 7Assemble JourneyTool’s AI drafts → you decide

Works with any chatbot — no setup, no connector.

In every path

  • You decide. Steps 1–4 and 6 count only once you press Confirm in the tool. In step 7 you can edit the teaching plan; your edits are saved automatically and go into the download.
  • Steps 5 and 7 always run inside the tool, and the number of AI generations per journey is limited.
  • What your own assistant or chatbot sees is governed by that assistant’s terms, not K8’s.
Start designing Privacy & cookies