We've all done RAG, now what?

In this episode of the Practical AI podcast, Daniel Whitenack interviews Rajiv Shah from Contextual AI about the advancements in AI, particularly in retrieval-augmented generation (RAG) and reasoning models. They discuss the common misconception of "training" AI models versus using retrieval methods to augment knowledge, and the importance of context engineering. Rajiv shares insights on navigating the complexities of AI implementation, highlighting the need to focus on practical use cases and organizational integration rather than just the latest technologies. The conversation also explores the challenges in scaling RAG systems, the evolving role of data scientists, and the potential of AI to assist in orchestrating traditional data science tools.