Contemporary Generative AI and Agentic Workflows
About this tutorial
This tutorial presents an introduction to the core concepts that underpin contemporary generative AI, together with key NLP methods that support many current agentic AI practices. The session covers large language models, prompt-based control, embeddings, retrieval-augmented generation, fine-tuning, and skill-based methods, with particular emphasis on how these approaches contribute to grounding model outputs, adapting model behavior to specialized tasks, and structuring reusable capabilities.
In addition to the conceptual material, the tutorial includes hands-on exercises in retrieval-augmented generation, fine-tuning, and skills, illustrating how these methods may be applied in practice and how they can be assessed in real-world AI systems.
Compute access
Frontier access is preferred, but a training account on Odo will be provided if needed.
What the session covers
Prompt-based control
Embeddings
Retrieval-augmented generation
Fine-tuning
Skill-based methods
Registration
This training event is now closed.
