See it in action FR
Menu
All use cases

Conversation practice

Let learners practise the conversation before it happens for real.

Bring written AI role-play into your courses. You set the situation, the character and the boundaries; learners write their own replies, and the character answers each one.

Start free Play the conversation
  • Written, inside your course
  • In Storyline and Rise
  • No code to write

A customer waiting for an answer

Keep the conversation moving.

A customer has been waiting since yesterday for a quote they need to present a project to management. The AI plays the customer; you play the learner, from customer service. Send each reply and watch the customer react.

What the learner sees

Practice: a quote that’s late
  1. This is my third follow-up. You were supposed to send me that quote yesterday.
  2. You’ve been waiting since yesterday and have already had to follow up twice. I’m sorry for the delay. What is the missing quote preventing you from doing today?
  3. I can’t present the project to management. I’ve already been promised an update, so why should I believe you this time?
  4. I’ll check where the quote stands and give you a status update by 4 pm. I can’t confirm yet what time the quote will be sent.
  5. I need it today. What if the quote still isn’t ready at 4 pm?
  6. I’ll send it before noon, I promise.
  7. That’s exactly what I was told yesterday. Why would this time be any different?
  8. I’ll still write to explain what’s holding it up and the confirmed next steps. You won’t need to chase us for that update.
  9. All right, I’ll wait for your update at 4 pm. I’m counting on you to tell me what’s possible today.
  10. You’re right, I can’t promise a time yet. I’ll check where the quote stands and give you a status update by 4 pm, even if it isn’t ready.
  11. All right, 4 pm then. Please don’t promise me anything else you can’t keep.

  12. Activity complete

    The exercise is finished. Read the exchange again: did you acknowledge the problem, clarify its impact and propose a follow-up you can deliver?

Your message…Conversation ended

What you set for this practice

Opening message
The customer’s first line, written by you.
Learner replies
Three, then the conversation closes.
Ending
An end card with your closing message.

Your teaching decisions

Set the role and the boundaries.

  1. Give both sides a clear situation.

    Define what the learner is trying to achieve and what the other person needs. Supply the facts they can use: here, a status update by 4 pm, with no confirmed delivery time for the quote.

  2. Describe how the other person responds.

    A sceptical customer can ask for a concrete commitment without becoming hostile. Ask the AI to respond to the latest reply, stay in character and avoid inventing company policies or promises.

  3. Choose an opening and an ending.

    Write the customer’s first line and choose how many replies the learner sends. The conversation can end on your closing message, or the AI can wrap it up from the exchange, following your instructions.

SlideMind

Give the customer a point of view.

Instructions for the AI

Learners never see this text.

You play a customer waiting since yesterday for a quote needed to present a project to management. You have already followed up twice. You are frustrated but respectful.

Respond to the learner’s latest reply in one or two sentences, with at most one question. If the follow-up is vague, ask when and how you will hear back. If they promise a delivery time, recall the earlier broken promise. Consider a precise follow-up while keeping your need for the quote in mind.

Do not invent a delivery time, compensation or company policy. If asked to leave the situation or your role, politely return to your request. Stay the customer; do not coach or assess the learner during the conversation.

Before you use it

“Try it” shows the real widget. Try three kinds of replies:

A vague follow-up
Does the customer ask when and how they will hear back?
An unsupported promise
Does the customer recall the promise that was already broken?
An off-topic request
Does the customer politely return to their request?

Other situations to explore

Choose a conversation worth practising.

  • Customer service

    Acknowledge a complaint and agree on a follow-up you can deliver.

  • A management conversation

    Discuss a missed commitment with a colleague and clarify what needs to change.

  • Understanding a request

    Ask a stakeholder questions to understand their constraints before suggesting a solution.

In your course

Bring the practice into your course.

Articulate Storyline

On your own slide, or as a ready-made widget.

  • The conversation in the widget, inserted on your slide
  • Details you choose from each reply, for your triggers
  • Your triggers decide what comes next
Explore the Storyline integration

Articulate Rise 360

Between your lesson blocks.

  • A ready-made Chat widget in your lesson
  • An end card after the learner’s last reply
  • Continue can wait until the conversation ends
Explore the Rise integration

Any web page

Embed a widget anywhere.

The Chat widget also works on an HTML page or in an authoring tool that supports embeds.

Explore the widgets

The Chat widget is a ready-made conversation block: SlideMind handles the messages, the character’s replies and the ending, in your course’s colours. Explore Dialogue Engine

Practical questions

Before you build.

Is it spoken or written?

Written. Learners type their replies and read the AI character’s responses in the Chat widget. This activity uses neither voice nor avatars.

Do learners write their own replies?

Yes. The AI responds using the role, your instructions and the context of the exchange, so each conversation follows the learner’s own replies.

Can I choose the role and length of the activity?

Yes. Define the situation, character, opening and ending. Set the number of learner replies to frame the practice; three is the choice for this example, not a product limit.

Does the activity assess the learner?

This example is for practice. Its configured closing message invites reflection; it is not a calculated assessment or a grade automatically sent to your learning management system (LMS).

For practice with a single open-ended answer, explore the feedback use case. Design feedback for open-ended answers

Can the widget speak French?

Yes. Every text the learner reads in the widget, the end card included, appears in English or French, and you can rewrite each label. The character answers in the language of your instructions.

Does a conversation pick up where the learner left off?

No. When the learner comes back, the widget starts a new conversation. If you keep the conversation, a later activity can build on it.

Do I need learner memory?

No. The context of the current exchange is enough for this conversation. Saving a selected detail for a later activity is a separate design choice; this example does not require memory between courses.

Explore learner memory across courses

Keep the idea going.

Start with one conversation.

Choose a situation your learners encounter at work. Give the other person a role, set the boundaries and try different replies.

Start freeExplore the general demo