Generative AI Solution Development with Databricks

In this session, you'll get an introduction to building contextual Generative AI applications using Retrieval-Augmented Generation (RAG) on Databricks. We'll walk through how Databricks supports each step of the RAG workflow: from embedding and indexing your data to generating more accurate and reliable AI responses using Mosaic AI and Vector Search.
You'll also learn how to prepare your own unstructured data, compute embeddings, and set up a vector search index in minutes. Whether you're just exploring Generative AI or already working with LLMs, this session will give you a practical foundation for deploying real-world GenAI solutions on the Databricks Data Intelligence Platform.

Topics
Generative AI promises a lot—but only works well when it has the right context. This session focuses on how to bridge the gap between generic LLMs and your specific data using RAG, a powerful architecture that brings your content into the conversation.
We'll simplify the RAG pipeline, clarify when and why to use vector databases, and show how to combine Databricks-native features like Delta Lake, Mosaic AI Model Serving, and Vector Search to build end-to-end GenAI solutions. Expect demos, clear architecture breakdowns, and actionable takeaways you can start using right away.
The session will include a live demo with:

• Retrieval-Augmented Generation (RAG)
• Show how model responses change when relevant context is added
• Use Mosaic AI Playground to simulate prompt hallucinations vs. grounded answers
• Vector Search with Databricks
• Create a vector index from document chunks
• Search for semantically similar content using a sample query

Is this for you? Are you a data wizard, ML tinker, or AI enthusiast? Whether you're a Data Scientist, ML Engineer, Data Engineer, or just someone who loves making LLMs smarter, this session is for you. If you're working with AI and want to supercharge your models with real-world data, don't miss it!

What Helps to Know: A basic understanding of NLP, prompt engineering, and Databricks will help—but if you’re curious and eager to learn, you're already halfway there!

📋Agenda
18:00 Registration & Welcome
18:30 Session and Q&A
19:30 Drinks & Bites
We look forward to meeting you! 🤝