AI x Design Workshop | 2nd Edition

This two‑day intensive training stands apart from all the typical AI courses in Mauritius. Instead of teaching cloud tools that lock you into subscriptions, it shows architects, researchers and students how to run their own AI models directly on their laptops — no internet, no external servers, no data leaving the room. Through hands‑on case studies with ComfyUI, Ollama and AnythingLLM, participants learn how to generate images, styles, atmospheres and technical assistants using fully local, open‑source systems they control from end to end. A bold, practical immersion for anyone who wants to understand how AI really works, build their own tools, and stop depending on commercial platforms.

International lecturers:

Prof. Laurent Lescop, architect and AI expert

Dr. Claudia Enrech, learning engineer

Ms. Divyaa Hauzaree, local AI specialist

Workshop Programme

19 Nov - 9.30 to 16.30

• Understand how LLMs work — tokens, context, hallucinations, open vs closed models, data confidentiality.

• Deploy your first local LLM — download, quantize, test and compare models on real agency tasks. “

• Code with a local model — generate and correct scripts directly from your local AI server.

• Master image generation with ComfyUI — prompts, image‑to‑image, ControlNet, LoRA, and full visual control.

• End‑of‑day synthesis — what runs locally vs what stays in the cloud, and how to plan your own tools.

20 Nov - 9.30 to 16.30

• Learn the vibe‑coding method — short iterations, quick tests, versioning, and choosing your mini‑project.

• Create your first AI app in the cloud — use Google AI Studio to generate and refine a working web tool.

• Move your project to local execution — use Claude Code to build, test, enrich, and version your tool.

• Go 100% local — run your tool with local models (Qwen, gpt‑oss, etc.) and compare performance.

• Final presentations — each participant shows their tool and explains which AI setup fits which task.

Conference

19 Nov - 18.00 to 19.30 : Building Digital Independence: Issues and Practices

This conference explores the misconceptions surrounding artificial intelligence, from exaggerated promises to misleading uses that fuel public confusion. It demystifies how AI actually works, its capabilities as well as its limits, and shows why many common narratives are technically inaccurate. Through simple demonstrations and concrete exchanges, participants learn to recognise widespread myths, identify situations where AI may produce unreliable results, and understand how certain uses or narratives can manipulate perception. The aim is to provide practical reference points that help everyone build genuine digital autonomy: understanding, verifying and using AI with clarity, critical thinking and informed confidence.

(Free Access Conference).

Technical Requirements

To ensure smooth participation, we recommend that each attendee has the setup listed below.

• A computer with at least 16 GB RAM (32 GB recommended)

• 50 GB free disk space

• For image generation: NVIDIA RTX GPU or Mac Apple Silicon M3–M4

• A Google account (for AI Studio access)

• Preferably an external SSD 1 TB for model storage

• Software and models downloaded before the workshop (several GB)