Creating a RAG app PoC is easy. Scaling it? Not so much.
In this session, we break down what it really takes to move from a quick RAG demo to a scalable, reliable system — without duct tape and fragile hacks.
You’ll learn:
- How to build a solid document ingestion pipeline
- Smarter retrieval strategies (hybrid, reranking, compression)
- Why observability matters (and how to set it up fast)
- What to consider when productizing your RAG stack
We’ll show practical examples using ragbits 🐰 — our open-source framework built for production-grade GenAI systems.
Plus, a sneak peek at upcoming Agentic features: tool use, memory, and custom UIs.
🔗 GitHub: https://github.com/deepsense-ai/ragbits
Timeline
00:00 Intro & Overview of RAG Systems
04:50 What is ragbits and Why We Created It
08:07 How to Build Successful RAGs
08:50 Ingestion Pipeline
16:07 Retrieval Strategy
24:47 Debugging & Observability
30:00 Next Releases
Speaker
Mateusz Hordyński
Senior Tech Lead at deepsense.ai
