KAI: AI assistant for e-learning courses in Knowledgeworker Create

KI-KAI is particularly suitable for teams that regularly create online courses and want to turn existing information into an editable course draft more quickly. 

Typical situations include: 

• an initial online course needs to be created from a PowerPoint presentation or PDF, 
• a mandatory module needs to be updated at short notice, 
• a topic needs to be prepared clearly for different target groups, 
• existing learning texts need to be shortened, expanded, or standardized in tone and language, 
• questions and feedback need to be created more quickly, 
• courses need to be translated into additional languages, 
• authoring teams want to use AI without leaving their central review and approval process. 

This final point is particularly important in organizations. AI only truly saves time when results do not then have to be laboriously transferred to another system and reformatted there. 

With the KI-KAI CourseWizard, you can develop an initial course draft from an idea. You specify the topic, target group, language, desired length, and tone of voice. KAI then suggests a title, description, outline, and content. This draft is not a finished course that should be published without review. 

It is a faster starting point for editorial and subject-matter work. Authors, subject-matter experts, and project managers can then adapt the structure, add content, and correct priorities. 

Many e-learning projects do not start from scratch. Presentations, manuals, training materials, or white papers are often already available. KI-KAI can use such source materials as a basis and prepare a more instructionally meaningful course structure from them. This is particularly helpful when existing content is valuable from a subject-matter perspective but has not yet been prepared for learning. 

A presentation is rarely a good online course on its own. KAI helps transform raw material into smaller learning units, clear text, and suitable course sections. 

In the course editor, KI-KAI supports the editing of existing text. Content can be shortened, expanded, simplified, or adapted linguistically. This makes it faster to create variants for different target groups or tones of voice.

The advantage lies in the context: the text remains part of the course. Authors can immediately see whether the length, style, and placement fit the learning page.  

Good knowledge-check questions need more than correct and incorrect answer options. They should test whether learners have understood the key message and provide meaningful feedback.

KI-KAI supports the creation of these types of questions. Existing learning content can be used to generate question suggestions with answer options and feedback. People then review whether the question is factually correct, fits the learning objective, and neither under-challenges nor overwhelms learners.  

KI-KAI can generate suitable images directly in the authoring process. This helps when a course should not consist only of text but no suitable images are available.

Quality control remains important here as well: Does the image fit the learning objective? Are the details correct? Is the style consistent? Especially in corporate training, visuals should support the content rather than merely decorate it.  

Text-to-speech can convert learning text directly into audio. This can make courses more accessible, support international target groups, or offer learners an additional way to engage with content.

Audio should not automatically duplicate every text. It is useful where spoken language provides orientation, reduces cognitive load, or makes the course more accessible.

KI-KAI supports the translation of courses into additional languages. This significantly reduces effort compared with manual export and import processes.

Nevertheless, every language version needs to be reviewed. Technical terminology, tone of voice, cultural examples, and layout must be checked. AI accelerates translation; responsibility for quality and approval remains with the team.  

General AI tools such as ChatGPT, Claude, or Gemini are useful for ideas, summaries, outlines, and text drafts. For individual tasks, they can provide strong preparatory work very quickly.

In e-learning production, however, much of the effort often begins after the first draft: content has to be transferred to course pages, formatted, reviewed, expanded, translated, versioned, and published. This is precisely where an AI assistant in an authoring tool differs from an external chat.

The question, therefore, is not whether general AI tools or KI-KAI are better. What matters is where the course will ultimately be created and maintained. External tools are useful for individual preparatory tasks. For scalable course creation, an integrated AI process is usually more efficient.  

KI-KAI accelerates production, but it does not replace responsibility. Especially for mandatory training, compliance topics, product training, or safety-critical content, people must verify that statements are correct, complete, and instructionally sound.

A strong AI-supported authoring process therefore includes defined review steps:

1. Define the learning objective and target group.
2. Select and approve the source material.
3. Generate a KAI draft.
4. Review the structure and content for subject-matter accuracy.
5. Refine the instructional design, clarity, and tone of voice.
6. Check questions, feedback, and media.
7. Test and approve the course technically.

This makes AI not a shortcut around quality, but an accelerator for a controlled production process.  

KI-KAI is especially worthwhile when your team regularly creates or updates learning content. The more often similar work steps recur, the greater the benefit:

• faster initial drafts,
• less manual transfer,
• more consistent language,
• faster creation of variants,
• more efficient translation,
• better use of existing materials,
• more time for subject-matter and instructional quality.

For a single small e-learning course, a general AI tool may be sufficient. However, when many courses, multiple authors, different languages, and recurring reviews come together, the integrated workflow becomes more important.