To RAG Or Not To RAG

Artificial intelligence is everywhere right now, and one of the most frequently recommended patterns for building AI-powered applications is Retrieval-Augmented Generation (RAG).

But what does that actually mean in practice? How do you implement it? What does it require? And perhaps most importantly, do you really need it?

Key takeaways

What RAG actually is and how it works What infrastructure and data are required to implement it When RAG is useful (and when it may be unnecessary)

Who Is This For?

• Software engineers exploring AI-powered applications
• Developers experimenting with LLM integrations
• Architects evaluating when RAG makes sense

Level

Practitioner

What This Session Covers

In this session, Piotr explores the practical side of Retrieval-Augmented Generation.

Instead of treating RAG as a buzzword or default architecture, the talk examines when it is useful, what it takes to implement it, and when it may not be the right solution.

Through working code examples, you’ll see how RAG systems are built and what trade-offs they introduce.

Talk style

• Code walkthrough with some editing / switching parts on and off
• Practical implementation discussion
• Real-world engineering considerations