Short learning units in the chat
Learning sequences with videos, interactive exercises and knowledge checks. Usable on mobile, right where your employees already ask their questions.
i40 / AI Learning Buddy
Your employees already use Copilot or an internal chatbot in their daily work. We build a learning buddy right into it: short learning units, knowledge checks and examples from your own documents. Technology, content and instructional design come from a single source.
Discuss your learning buddyAnimated example: an employee learns with the Learning Buddy in the Microsoft 365 Copilot chat and answers a knowledge check. The Practice Companion then checks their draft against the sources.
Learning sequences with videos, interactive exercises and knowledge checks. Usable on mobile, right where your employees already ask their questions.
Role-specific examples from production, maintenance, quality, procurement and engineering, based on your real documents rather than made-up case studies.
Access to company data runs through RAG and each user's existing access rights. No blanket service account.
We keep content structured. Adding another language does not mean producing it again.
Practice Companion
Between training and live sessions, each target group practises on its own workflow. The Practice Companion sets practical tasks with illustrative data, checks drafts against the sources and points to approved workflows.
Data & operations
Running the buddy and storing data in your own infrastructure is our standard case. We provide the content, learning logic and configuration.
If you prefer, we operate the learning buddy ourselves, hosted in the EU.
In developing agents and accessing company data via RAG.
Sample questions from the learning path “AI literacy in operations”
| Role | Sample question in the chat |
|---|---|
| Production | How do I check whether an AI forecast of the scrap rate can be right? |
| Maintenance | What does an 80 per cent probability of failure mean for my planning? |
| Quality | What data does image recognition need to detect defects reliably? |
| Procurement | How do I write a good prompt for a supplier comparison? |
| Engineering | When is an AI use case worth it, and when not? |
As an agent in your company's Copilot environment.
On Azure, for example, as a learning feature in your existing assistant.
On request. We work out with you which way of embedding it fits best.
| Area | i40 | Your company |
|---|---|---|
| Content & instructional design | Learning sequences, videos, exercises, knowledge checks, role-specific examples | Subject-matter expertise and documents |
| Technology | Learning logic and configuration of the learning buddy | Your assistant: Copilot or your own chatbot |
| Data access | RAG via each user's existing access rights | Access management as before |
| Operation | Optional, then in the EU | Standard case: in your infrastructure |
Which assistant, which data sources, which target groups.
Define learning paths, role examples and knowledge checks.
Test with one target group and refine.
Add further roles and languages.
Our foundation
With SkillDuck®, we develop and operate an AI-powered authoring platform for microlearning. That experience goes into the learning buddy.
Explore SkillDuck®Contact
Let's talk about your AI environment, your target groups and the right way to get started.