22title : " How to Build a Real-Time AI RAG Pipeline"
33---
44
5+ import ReactPlayer from ' react-player' ;
6+
57In this guide, we will build a complete, end-to-end Retrieval-Augmented
68Generation (RAG) system using Conduit. This powerful pipeline will automatically
79ingest new documents, process them into a queryable format, and make them
@@ -20,14 +22,12 @@ This first part introduces the problem we aim to solve and gives a high-level
2022overview of the solution's architecture.
2123
2224<div style = { { position: ' relative' , paddingBottom: ' 56.25%' , height: 0 , overflow: ' hidden' , maxWidth: ' 100%' }} >
23- <iframe
24- src = " https://www.loom.com/embed/855eeea8cf2e439c81650dcb5c16277e?sid=6974e996-8e73-4fa0-b2fc-1084062627a3?hideEmbedTopBar=true&hide_title=true&hide_owner=true&hide_share=true&hide_speed=true"
25- webkitallowfullscreen
26- frameborder = " 0"
27- mozallowfullscreen
28- allowFullScreen
29- style = { { position: ' absolute' , top: 0 , left: 0 , width: ' 100%' , height: ' 100%' }}
30- ></iframe >
25+ <ReactPlayer
26+ style = { { position: ' absolute' , top: 0 , left: 0 , width: ' 100%' , height: ' 100%' }}
27+ controls
28+ playIcon
29+ src = ' /video/rag-guide-01-building-an-ai-pipeline.mp4'
30+ />
3131</div >
3232
3333### The Problem
@@ -107,14 +107,12 @@ raw data.
107107![ Architecture Diagram] ( /img/guide-rag-ollama-architecture-part2.png )
108108
109109<div style = { { position: ' relative' , paddingBottom: ' 56.25%' , height: 0 , overflow: ' hidden' , maxWidth: ' 100%' }} >
110- <iframe
111- src = " https://www.loom.com/embed/d213d448ebc349ff8e5a22d7884c2f30?sid=1855d6ef-7cea-4525-a7eb-d151f198750c?hideEmbedTopBar=true&hide_title=true&hide_owner=true&hide_share=true&hide_speed=true"
112- webkitallowfullscreen
113- frameborder = " 0"
114- mozallowfullscreen
115- allowFullScreen
116- style = { { position: ' absolute' , top: 0 , left: 0 , width: ' 100%' , height: ' 100%' }}
117- ></iframe >
110+ <ReactPlayer
111+ style = { { position: ' absolute' , top: 0 , left: 0 , width: ' 100%' , height: ' 100%' }}
112+ controls
113+ playIcon
114+ src = ' /video/rag-guide-02-reading-from-s3.mp4'
115+ />
118116</div >
119117
120118### 1: Prerequisites
@@ -266,14 +264,12 @@ processors to partition the PDF content into smaller chunks using the
266264
267265
268266<div style={{ position : ' relative' , paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%' }}>
269- <iframe
270- src="https://www.loom.com/embed/7065a048cbde4a43925c6528f6dec1a9?sid=1dd3b5f4-a5d2-47e5-8401-b1670a2a4977?hideEmbedTopBar=true&hide_title=true&hide_owner=true&hide_share=true&hide_speed=true"
271- webkitallowfullscreen
272- frameborder="0"
273- mozallowfullscreen
274- allowFullScreen
275- style={{ position : ' absolute' , top: 0, left: 0, width: '100%', height: '100%' }}
276- ></iframe>
267+ <ReactPlayer
268+ style={{ position : ' absolute' , top: 0, left: 0, width: '100%', height: '100%' }}
269+ controls
270+ playIcon
271+ src='/video/rag-guide-03-chunking-documents.mp4'
272+ />
277273</div>
278274
279275# ## 1: Set Up the Unstructured.io Service
@@ -357,7 +353,14 @@ text.
357353
358354
359355
360- [TODO : VIDEO PLACEHOLDER FOR PART 4]
356+ <div style={{ position : ' relative' , paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%' }}>
357+ <ReactPlayer
358+ style={{ position : ' absolute' , top: 0, left: 0, width: '100%', height: '100%' }}
359+ controls
360+ playIcon
361+ src='/video/rag-guide-04-creating-embeddings.mp4'
362+ />
363+ </div>
361364
362365# ## 1: Set Up Ollama
363366
@@ -481,4 +484,4 @@ available to our chatbot UI.
481484
482485[TODO : VIDEO PLACEHOLDER FOR PART 5]
483486
484- TODO : text
487+ TODO : text
0 commit comments