Development of an AI-enhanced microlearning platform for enterprise knowledge transfer

In the modern enterprise environment, continuous training is crucial but often hampered by expensive and inflexible traditional methods. Microlearning, an approach that delivers training content in short, focused modules, is emerging as an effective solution to improve engagement and knowledge retention. However, the manual creation of high-quality microlearning content is a significant barrier, requiring time and instructional design skills. This thesis addresses this challenge through the development of an innovative platform enhanced with Artificial Intelligence, aimed at automating and simplifying knowledge transfer within the enterprise. The main objective is to design and implement a web application that enables users, particularly Subject Matter Experts, to rapidly transform raw corporate documents (such as PDF, DOCX, XLSX, TXT files, and URLs) into interactive microlearning modules. The methodology is based on a modern and scalable software architecture. The backend, developed in Python with FastAPI, orchestrates complex workflows using LangChain and LangGraph. The frontend, built with React and Next.js, provides a responsive and intuitive user interface, enriched by a chatbot assistant powered by CopilotKit. The platform's intelligence is fueled by state-of-the-art AI models: GPT-4.1 for text generation and summarization, DALL·E 3 for creating contextual images, and the Tavily API for online resource retrieval. The entire solution is deployed on a serverless AWS infrastructure, utilizing services like Lambda, ECR, CloudFront, CloudWatch, CloudFormation, and Secret Manager to ensure scalability and security. The result is a fully functional platform that automates the content creation pipeline: from analyzing the source document to generating concise slides and interactive quizzes. Users can customize many aspect of the generated content through an assisted editor and receive real-time support from the chatbot. In conclusion, this work demonstrates the feasibility and effectiveness of an AI-driven approach for producing corporate training material. The developed platform provides a practical contribution by drastically reducing the time and costs associated with content creation, making microlearning a more accessible and scalable strategy for organizations.