See it in action FR
Menu

AI interactions in your courses

Dialogue Engine

AI questions and conversations

Invite learners to explain a decision or practise a conversation. AI reads what they write and responds according to your instructions: feedback on an open answer with Q&A, a character who reacts with Chat.

Start with one activity in your own course.

The course and SlideMind exchange a connected series of messages.
Dialogue Engine

A look inside

Discover Dialogue Engine

How it responds

What the engine receives, and what it returns.

Set by you in SlideMind

  • Your instructions

    The objective, criteria and tone; for a conversation, the role, the facts and when it ends.

  • The context

    Information set for the course and module, values sent by your course and, with Memory Engine, a detail saved earlier.

Written by the learner in your course

  • The learner’s answer

    Their own words, or the conversation so far.

Dialogue Engine

The AI model you choose applies your instructions.

  • A reply for the learner

    Feedback on the answer, or the character’s next line.

  • Values for your course

    Optional: a score, a pass or fail, or another field you define, to use in Storyline or show in the Q&A widget.

The engine does not decide how this appears. Your authoring tool or a ready-made widget handles the display.

Q&A · Workplace health and safety

One question, three answers.

The engine reads each answer against your instructions. It recognises what works and points to what is missing, without giving away the solution.

Design feedback for open-ended answers

Question

A colleague removes a machine guard to work faster. What do you do?

Instructions for the AI

In SlideMind. Learners never see this text.

You are a workplace health and safety trainer. The learner sees a colleague remove a machine guard to work faster. A good answer first has the machine stopped, protects the colleague without blaming them and reports the situation to a supervisor. Point out what the learner did well, then guide them towards what is missing with a question. Don’t give the full answer. Be direct and respectful, in three sentences at most.

  1. Partial answer

    I ask them to put the guard back.

    Feedback

    Good reflex: you step in straight away. Think about what comes next too: the machine must be stopped while the guard goes back on, and who should hear about what happened?

  2. Complete answer

    I ask them to stop the machine, we put the guard back together and I tell our supervisor.

    Feedback

    You cover the essentials: stop, secure and report. Reporting also helps understand why your colleague wanted to go faster.

  3. Off-topic answer

    These machines are too old anyway.

    Feedback

    The state of the equipment can be a real issue, and worth reporting too. But right now your colleague is exposed to a hazard: what do you do first?

Chat · Management

A character who reacts to the learner.

Describe the character, what they know and how they behave. AI plays the role and adjusts each reply to the way the learner leads the conversation.

Design a conversation for learners to practise

Instructions for the AI

In SlideMind. Learners never see this text.

You play Alex, a technician who has arrived late several times over the past two weeks. The learner is your manager. If they open with blame, stay defensive. If they ask an open question, explain that your daughter’s new daycare doesn’t open until 8:00. The team may start between 7:00 and 9:00 with the manager’s agreement: accept a solution that respects this rule. Answer naturally, in two or three sentences.

Conversation settings

The learner can send 3 messages. After the last one, the AI wraps up.

What the AI does on the last turn: ask the learner when Alex started to open up, and why.

  1. Alex · AI

    You wanted to see me?

  2. Learner

    Yes. You’ve arrived late several times over the past two weeks.

  3. Alex · AI

    I know. I make up the time in the evening, so it’s not a big deal.

    Defensive: the conversation opened with blame.

  4. Learner

    I can see you’re making an effort. What has changed recently?

  5. Alex · AI

    My daughter started at a new daycare. It doesn’t open until 8:00, so I can’t get here before 8:30.

    Opens up after an open question.

  6. Learner

    Thank you for telling me. Could we shift your hours, starting at 8:30 and finishing later?

  7. Alex · AI

    Yes, that would make things much easier. Thank you.

    Accepts a solution the scenario allows.

AI wrap-up

Practice complete. Looking back, when did Alex start to open up, and why?

Your teaching decisions

You set the frame. AI responds to the learner.

  1. Define the objective.

    Choose what learners should explain or practise. Select Q&A or Chat to suit the activity.

  2. Set the instructions and context.

    Criteria and tone for Q&A; a role, situation and boundaries for Chat. Supply the facts the AI needs to respond.

  3. Try it and refine.

    Use “Try it” with a partial answer, a complete answer and an off-topic response. Refine the instructions based on what you observe.

From intention to integration

Bring the interaction into your course.

Configure the activity in SlideMind, then choose how it appears in your course.

In Storyline, on your own slides

Keep your own slide design: the learner answers in your course, and the feedback appears where you choose. A score or a pass or fail you define can drive your triggers and branching.

Explore the Storyline integration

Anywhere, with a ready-made widget

Paste the Q&A or Chat widget into a Rise Code block, an HTML page or an authoring tool that supports embeds. The widget handles the display: question, answer and feedback, with an optional score and verdict for Q&A.

Explore the widgets

Test the activity in your published course. Each integration page explains its setup and completion options.

Practical questions

Before you build.

Can I choose what the AI takes into account?

Yes. Set the context, criteria, role and tone for your activity. Your instructions guide the response; try several learner answers to see where the instructions need refining.

Do I need Memory Engine?

No. Your instructions and the current activity’s context are enough. Memory Engine can add a selected detail saved earlier. Saving information for another activity is separate from the conversation in progress.

Does the result become a grade in my learning management system (LMS)?

A Q&A score displayed in the widget is separate from an LMS grade. In Rise, SlideMind passes only the completion of the activity, not the score. In Storyline, returned values feed the variables and rules you configure in your course.

Do I need to write code?

SlideMind generates the integration code. Paste it into a Rise Code block, or connect variables and triggers in Storyline. In Storyline, the Q&A and Chat widgets also fit in a Web Object. The integration pages walk through each approach.

Keep the idea going.

Start with a question or a conversation.

Choose an activity in your course, set the instructions and explore the responses your learners could receive.