AI can do more than answer questions in a chat window. In this hands-on workshop you will discover what happens when AI agents are given a goal, tools and context, and start carrying out complete steps of your work — from formulating learning objectives to producing educational media and adapting the learning offer to each individual student.
Workshop leader Serge de Beer develops agentic systems for schools, universities and media organisations, and leads the Agentic Production programme for the Dutch broadcasting industry. In this deep dive he brings that practice to the Media & Learning community: no slides full of promises, but live systems, honest failures and a method you can apply to your own work.
What we will do during the three hours:
Part 1 — Understanding agents (about 45 minutes)
What makes a system “agentic”, and how does it differ from chatting with AI or classic automation? Using the anatomy of an agent — a goal, tools, context and a control loop — we map the full Agentic Learning chain: from learning objectives, through the design and production of learning materials, to learning analytics and a hyper-personalised offer for each student.
Part 2 — Agentic Production live (about 60 minutes)
We follow one learning objective through a working production pipeline: an agent turns it into a script, generates images, voice-over and music, assembles a rough cut and prepares publication — with human checkpoints at the moments that matter. You will see exactly where agents save serious time in pre-production, production and post-production, and where human judgement remains indispensable.
Part 3 — Your own agentic workflow (about 60 minutes)
In small breakout groups you take one process from your own practice — a recurring production task, a support workflow, a piece of learning design — and design an agentic version of it: which goal, which tools, which context, and where you keep a human in the loop. Serge supports the groups individually, and we close by sharing designs and discussing a realistic first step you can take at home.
After this workshop you will be able to:
– explain what agentic AI is and recognise which parts of your own workflow are suited to it;
– describe what agents can (and cannot yet) do in each phase of educational media production;
– place media production within the wider Agentic Learning chain, from learning objectives to personalised learning;
– take home a first design of an agentic workflow for one of your own processes.
Who is it for?
Professionals in educational media and learning support: media and video teams, instructional designers, learning technologists and managers who want to look beyond the chat window. No programming experience is needed.
Preparation
Bring one process from your own practice that you would love to hand over to an agent — something you do every production or every course, again and again. That case will be your working material in part 3.
Workshop leader: Serge de Beer, LearningTour, the Netherlands
Due to the workshop format, places are limited to 30 participants. Register early to secure your spot.
This workshop is free for MLA members!
