Conference Opening by MWS Chairs

Keynote: Scaling Live Streaming for the World

Vibe Coding: Superpower or Footgun? A Hands-On Reality Check

So, How Are Your Agents? — 365 More Days in the Life with AI

Voice AI for the professional Media Post-Production

Reinventing Content Discovery with Generative AI: Automated Previews, Chapters, and Highlights at Scale

Beyond Dashboards: How ARTE and EINBLIQ.IO Use AI Agents for Automating Streaming Operations

AI in Ad Sales: How RTL Uses AI Across the Entire Sales Funnel

The agentic AI mindset

Memes, Internet Culture & AI: How We Build Automated Workflows for Viral Social Media Content

Wikidata and Journalism in the Age of GenAI

Building on the foundational understanding from the first tutorial, this session focuses on how to develop AI‑powered applications tailored to the media domain. It covers hands‑on methods, architectural approaches, and best practices for building robust, production‑ready AI systems.

Participants learn how to work effectively with LLMs — from prompting and orchestration to fine‑tuning and advanced retrieval techniques (RAG). We explore how to integrate multimodal models into production workflows, build intelligent assistants, and design agent‑based pipelines capable of handling complex tasks.

Topics include:

• LLM-Based Application Development: Prompting strategies, evaluation, orchestration
• Fine‑Tuning & Customization: Parameter‑efficient methods, domain adaptation
• Retrieval‑Augmented Generation (RAG): Document pipelines, vector databases, retrieval workflows
• AI Agents: Tool use, multi‑step reasoning, automation patterns
• Multimodal AI: Vision‑language models, video analysis, interactive media chatbots
• Hands‑On Use Cases:
  • Intelligent document‑based chatbots
  • Search & recommendation assistants
  • Video‑based conversational agents
  • Interview bots for expert knowledge capture
• Authenticity & Security: C2PA, provenance, compliance, responsible AI guidelines

This tutorial is designed for developers, media technologists, innovation leads, and anyone creating AI-driven applications in media environments.

As AI becomes increasingly central to media workflows, understanding how modern models actually work is essential. This tutorial provides a clear, accessible, and technically grounded introduction to contemporary AI systems — with a special focus on neural networks, Transformer models, and Large Language Models (LLMs).

We unpack the mechanics behind today’s generative AI: how models process language, how attention works, what happens during training and inference, and why Transformers are so effective. Through intuitive explanations, visual breakdowns, and simplified examples, participants will gain a solid conceptual understanding of how modern AI systems operate “under the hood.”

Topics include:

• AI & Machine Learning Basics: Core concepts, model families, training principles
• Vectorization: Architecture, embeddings, tokenization, optimization
• The Transformer Model: Attention mechanisms, encoder/decoder structures, masking
• Capabilities & Limitations: Bias, hallucinations, safety constraints, model weaknesses
• Looking Ahead: Trends such as multimodal models and mixture of experts

This tutorial is ideal for participants who want to understand how modern AI and LLMs function.

The Agent-Ready Organization: Orchestrating Talent Velocity & Capability Sovereignty(in English)

Learning on Track – Lerntechnologien, zentral eingeführt – lokal wirksam

Paneldiskussion “Verändern wir das Lernen oder verändern die Technologien uns? – Über Chancen und Herausforderungen innovativer Lerntechnologien”

Symposium Closing Remarks

Symposium Opening

Frontier agentic Learning Experiences (Microsoft Perspective)

AI in Learning Management Systems

AI & Interoperability in School Education: The German Federal Approach Towards an Educational Ecosystem

Die digitale Bildungsplattform SCHULE@BW: aktueller Stand und Perspektiven.