RAGtime with Haystack! How to build a RAG with Haystack 2.0?
summary
RAGtime is a stream‑based demo that implements a Retrieval‑Augmented Generation pipeline using Haystack 2.0, uploads PDFs and text, converts and chunks then embeds documents into the Quadrant vector store, and calls the Hugging Face API for embeddings and generation. Pipelines shown: document conversion+preprocessing and retrieval+prompt‑builder, with modular features like reranking, compression, routing and self‑reflection. Demo runs with minor hiccups; production costs are order‑of‑magnitude hundreds of dollars per year for modest document volumes. Chunk sizes remain empirical.
AI-written summary · about
chapters (15)
- 0:00Introduction to RAG
- 0:37Why Use RAG?
- 3:31Core Components of RAG
- 6:00Advanced RAG Techniques
- 7:15Modular RAG Architecture
- 8:45RAG vs Fine-tuning
- 10:27Evaluation Metrics
- 11:55Introduction to Haystack
- 13:52Building Pipelines
- 14:55Demo: RAG Time App
- 17:10Demo Breakdown
- 18:40Pipeline Architecture
- 21:45Future Improvements
- 23:35Resources for Further Learning
- 24:52Q&A Session
related talks
42mTo RAG or not to RAG
21mHow RAG is Redefining AI Capabilities
39mRAGs to Riches: How Our Content Affects Retrieval Augmented Generation
01h 12mBuild RAG using Llama 3.1
01h 28mBuilding a Simple RAG Application: From Fundamentals to Implementation
01h 16mForget RAG Pipelines—Build Production Ready Agents in 15 Mins
39mWhat if you could build production-ready RAG entirely in Elixir?
32mRag Demystified