SkillDuck – AI Authoring Tool for E-Learning

Starting point

Creating high-quality e-learning courses at an EdTech company took weeks to months — from research and structuring to defining learning objectives and filling in content. Standard AI generation produced incoherent content with too many hallucinations.

Solution

Development of a multi-agent system that automatically generates complete, didactically structured e-learning courses from documents and general knowledge. Each agent handles a specialization: outline, learning objectives, narrative, didactic content. Human-in-the-loop cycles ensure quality at critical checkpoints.

Technology stack

• Azure AI Foundry, OpenAI GPT-4o (LLM Orchestration)
• Python, LangChain (Backend / Agent Framework)
• RAG architecture with vector indexing (context-sensitive generation from PDFs)
• Next.js, ReactJS (Frontend / Authoring Interface)
• Human-in-the-Loop Workflows (Quality Assurance / Hallucination Reduction)

Result & Impact

• Course creation time reduced from months to hours or days
• eLearning Award 2025 won in the AI category (client: i40 – the future skills company)
• Scalable to any document base and language