This demo-driven session delves into advanced aspects of Generative AI beyond basic prompt engineering. You’ll learn how GenAI works and how to implement use cases like classification, recommendation, and semantic search.

The session begins with core concepts such as tokens, exploring how transformers learn associations, followed by text generation via completions for a deeper understanding of model functionality. Key topics include embedding vectors, cosine similarity, and vector databases to explain semantic modeling.

You’ll also discover techniques like Retrieval Augmented Generation (RAG) to expand and manipulate model knowledge.

Finally, we’ll explore various model types and tools, including OpenAI Playground, Azure AI Studio, and Microsoft’s Semantic Kernel.

This technical session is ideal for those looking to understand and build intelligent applications powered by large language models.