LehrPuls

Live polling in the lecture hall — local, no cloud.

← hollomey.com Deutsch

Guides › Creating questions with AI

Creating questions with AI

With an AI service of your choice — such as ChatGPT or Claude — you have finished questions generated from your own material and bring them into LehrPuls in one go. LehrPuls itself contains no artificial intelligence and sends nothing to the internet: you hold the conversation in the service’s window, LehrPuls provides the instructions and checks the result. This guide goes through both routes with the bundled sample result.

A question collection on one topic

From the dialog via the check report to the imported questions in the topic tree.

  1. 1

    Open the dialog

    Step 1 of 8

    Open Topics & Questions and click AI above the question list — right next to Question. The Create questions with AI dialog opens.

    Screenshot Cockpit 11Topics & Questions22AI
    Cockpit
  2. 2

    What is to be created

    Step 2 of 8

    At the very top you choose what the AI service is to deliver. Question collection on one topic produces a handful of questions on one subject — for a chapter, an exercise or a review. Course unit with playlist turns structured lecture material into a whole unit including a playlist (second variant of this guide). Introduction, steps and sample prompt adapt; LehrPuls remembers your choice.

    Screenshot Cockpit 11Question collection
    Cockpit
  3. 3

    Five steps and the structure description

    Step 3 of 8

    The dialog guides you through the process in five steps. First: Save structure description (PDF) stores the format description as a file; View opens it for reading. You pass it on to the AI service unchanged — it needs nothing more. Second, you upload your source material and this file to the AI service. A sample lecture is included for trying it out: Save sample lecture notes (PDF) stores it, Insert sample result below puts the matching result right into the field below.

    Screenshot Cockpit 11the five steps22Save structure description33View structure description44Save sample lecture notes55View sample lecture notes66Insert sample result
    Cockpit
  4. 4

    Copy the prompt

    Step 4 of 8

    Third: below that is the sample prompt — the instructions for the AI service. Copy prompt puts it on the clipboard. Paste it into the AI service, adjust the number of questions, question types, times and difficulty to your liking, and send it. If you maintain your questions in several languages, the prompt asks for the other language versions right away.

    Screenshot Cockpit 11Sample prompt22Copy prompt
    Cockpit
  5. 5

    Paste the answer and read the check report

    Step 5 of 8

    Fourth: copy the AI service’s answer in full and paste it into the field Paste the AI's answer here at the bottom — here Insert sample result below has done that. LehrPuls checks the text by itself. The green line All good states the scope, below it every question is listed with title and type. If something is wrong, the reason is shown next to the question, Import stays locked, and Copy findings puts all findings on the clipboard for the AI service. Clear input empties the field for the next version.

    Screenshot Cockpit 11Result of the check22every question with type33The AI’s answer44Clear input55Import
    Cockpit
  6. 6

    Import and choose the target topic

    Step 6 of 8

    Fifth: if nothing is flagged, click Import and choose under Target topic where the delivery is attached. — Top level — creates the delivered topics as new main topics; an existing topic takes them in as subtopics. With Import .lpx you bring in a saved answer (.json) or an .lpx file.

    Screenshot Cockpit 11Target topic22Import
    Cockpit
  7. 7

    The questions in the topic tree

    Step 7 of 8

    The question collection is now in the tree: a topic with the title of the lecture notes, one subtopic per chapter below it, plus a scenario. Because it is the unchanged sample result, the topic carries the Sample badge and disappears again with Settings → Samples, Remove samples. A real AI answer becomes an ordinary topic.

    Screenshot Cockpit 11imported topic22one subtopic per chapter33Sample44the questions of the topic
    Cockpit
  8. 8

    Check every question

    Step 8 of 8

    An AI service writes clean language but not necessarily correct content. Read every question in the preview before using it for the first time, check especially the answers marked as correct, the wrong answer options and the model answers, and touch them up with the pencil in the question editor. This route does not transfer images — create image questions in the question editor.

    Screenshot Cockpit 11selected question22Edit in the question editor33Preview
    Cockpit

A course unit with playlist

Lecture material with chapters becomes topics, questions, a scenario and a ready playlist with a group exercise.

  1. 1

    Choose Course unit with playlist

    Step 1 of 7

    Under Topics & Questions, open the dialog with AI and choose Course unit with playlist at the top. This prompt turns material that is divided into chapters into a whole unit: a topic with the title of the course, one subtopic per chapter with four to six questions of mixed question types, model answer and time — plus a playlist.

    Screenshot Cockpit 11Course unit
    Cockpit
  2. 2

    The same five steps

    Step 2 of 7

    The introduction and steps now describe the course unit; the process stays the same: save the structure description, upload it to the AI service together with your lecture notes, copy and adjust the sample prompt, paste the answer below, check and import. To try it out, Insert sample result below inserts the result an AI service should produce from the sample lecture notes — this time with a playlist.

    Screenshot Cockpit 11the five steps22Save structure description33Insert sample result
    Cockpit
  3. 3

    The check report also names the playlist

    Step 3 of 7

    As with the question collection, the check runs by itself. The green line All good now names the playlist besides topics, questions and scenario; below it every question is listed with type and chapter. Import is available as soon as nothing is flagged.

    Screenshot Cockpit 11Scope with playlist22Questions with type and chapter33Import
    Cockpit
  4. 4

    Import and choose the target topic

    Step 4 of 7

    Click Import and choose the Target topic. With — Top level — the course unit becomes a main topic of its own; the playlist is created along the way.

    Screenshot Cockpit 11Target topic22Import
    Cockpit
  5. 5

    One topic per chapter

    Step 5 of 7

    The tree shows the course unit as a topic with one subtopic per chapter. The scenario for the group exercise — a realistic case spanning several chapters — belongs to the top topic and appears in the bar above the question list.

    Screenshot Cockpit 11Course unit22Subtopic per chapter33Scenario
    Cockpit
  6. 6

    The playlist is ready

    Step 6 of 7

    On the Playlists page, the new playlist is at the top. Its tree has one section per chapter and, as the last section, the group exercise. You can start it as it is or rework it like any other playlist.

    Screenshot Cockpit 11new playlist22Sections per chapter
    Cockpit
  7. 7

    Categories, peer instruction and group exercise

    Step 7 of 7

    Every question carries a category: ★ Must for the core concepts, ☆ Should for application and calculation, – Could for going deeper. In each chapter one question is set up for peer instruction (red ↻2). The group exercise at the end is a question block (⧉) with the scenario and several tasks; in the session you start it as group work via Arrange. The purple times show how long each section is planned to take.

    Screenshot Cockpit 11Category ★ ☆ –22Peer instruction33Question block of the group exercise44planned time
    Cockpit