CF

AI Engineering Bootcamp: RAG (Retrieval Augmented Generation) for LLMs

17h 51m 59s
English
Paid

This course shows you how to build smarter AI apps with RAG. You use RAG to give LLMs fresh facts from your own data. This helps you make tools like chatbots, search apps, and simple analysis systems.

Why RAG matters

LLMs learn from old training data. This can limit their answers. RAG helps you fix this by pulling new facts from files, links, and other sources. You then pass these facts to the model. This leads to clear and useful results in real tasks.

Example: A shop chatbot can check live stock data. It does not guess based on old samples. It gives you clear info about items and delivery dates.

What you learn

Retrieval basics

  • How to clean and prepare text for search
  • How Boolean, vector, and simple probabilistic search work
  • How indexing, queries, and ranking work

Generative model basics

  • How transformers and attention work
  • How to prepare data and train text models

Intro to RAG

  • How search and generation work together
  • How RAG helps in real projects

Using the OpenAI API

  • How to set up the API and write clear prompts
  • How model settings change output

Building RAG with OpenAI

  • How to build a simple end‑to‑end RAG system
  • How to link search results with model output

Working with unstructured data

  • How to read data from PDF, Word, PowerPoint, Excel, and images
  • How to pull useful text from mixed formats

Multimodal RAG

  • How to use both text and images in one flow
  • How to merge data types into one answer

Agent systems with RAG

  • How to build simple agents that talk to users and run tasks
  • How to track agent state and run steps in order

Why this course helps you

You get hands-on practice with key RAG steps. You learn how to build tools that use fresh data, handle wide search tasks, and give clear answers. These skills help you build strong AI apps for real work.

About the Author: Zero To Mastery

Zero To Mastery thumbnail

Zero To Mastery (ZTM) is a Toronto-based online coding academy founded by Andrei Neagoie, originally a senior developer at large Canadian tech firms before turning to teaching full-time. The academy's signature is the cohort-based bootcamp track combined with a deep self-paced course library, all aimed at career-changers and self-taught developers preparing to land software-engineering roles at top companies.

The instructor roster has grown well beyond Andrei to include other senior practitioners: Daniel Bourke (machine learning), Aleksa Tešić (DevOps), Jacinto Wong, and others. Courses cover the full software-engineering career path: web development with React and Next.js, Python, machine learning and deep learning, DevOps and cloud, system design, mobile, and the algorithm / data-structure interview prep that gates engineering jobs.

The CourseFlix listing under this source carries over 120 ZTM courses spanning that full range. Material is paid; ZTM itself runs on a monthly / annual membership model. The teaching style favours long-form, project-based courses where students build complete portfolio-quality applications rather than disconnected feature tutorials.

Watch Online 171 lessons

This is a demo lesson (10:00 remaining)

You can watch up to 10 minutes for free. Subscribe to unlock all 171 lessons in this course and access 10,000+ hours of premium content across all courses.

View Pricing
0:00
/
#1: Course Outline
All Course Lessons (171)
#Lesson TitleDurationAccess
1
Course Outline Demo
03:11
2
Meet Rubber Ducky! Your AI Course Assistant using RAG
03:11
3
Who Is This Part For?
02:16
4
Game Plan for Prompt Engineering Basics
04:31
5
Setting Up the OpenAI API
03:04
6
Few-Shot Prompting
03:11
7
Few-Shot in Practice
09:54
8
Role, Persona and Goal
04:59
9
Role, Persona and Goal in Practice
04:45
10
System Message
05:05
11
System Message in Practice
06:02
12
My Favourite Prompt
04:03
13
Understanding Transformers
12:48
14
Attention Mechanisms
06:15
15
Game Plan for Python for RAG and GenAI
01:35
16
Loops
05:19
17
Loops: Easy Level
08:32
18
Loops: Medium Level - Part 1
03:46
19
Loops: Medium Level - Part 2
03:56
20
Loops: Hard Level
02:57
21
Functions
04:44
22
Functions: Easy Level - Part 1
04:07
23
Functions: Easy Level - Part 2
01:31
24
Functions: Medium Level - Part 1
02:50
25
Functions: Medium Level - Part 2
03:08
26
Functions: Hard Level
07:01
27
Introduction to Classes
04:52
28
Classes: Easy Level - Part 1
10:30
29
Classes: Easy Level - Part 2
03:53
30
Classes: Medium Level
08:43
31
OpenAI Tokenizer
06:13
32
Overview: Working with the OpenAI API
03:48
33
OpenAI API for Text
04:53
34
Setting Up OpenAI API Key
05:09
35
OpenAI API
05:02
36
Generating Text with OpenAI API
06:38
37
OpenAI API Parameters
06:55
38
OpenAI API for Images
04:52
39
With Image URL
09:20
40
With Image in Base64
10:09
41
Adding Few-Shot Prompting
06:27
42
What Did You Learn in this Section?
03:51
43
Playing the Dice, Rock, Paper, Scissors, and Guess the Number
08:12
44
Project Presentation: Build a LinkedIn Post Writer App
02:58
45
UI Design via Image Generation
07:46
46
Lovable Build Prompt
05:30
47
Deploy on Lovable
11:51
48
What to Expect of Part B
04:33
49
OpenAI File Search
02:05
50
Project Presentation: Build a Mini Rubber Ducky
03:44
51
Vector Stores
04:06
52
Setup
01:57
53
Retrieving the Files Path
06:55
54
File and Vector Stores in OpenAI
09:53
55
Responses Endpoint with File Search
09:33
56
Setting Up on Cursor and Requirements
05:36
57
Building Your AI Web App
02:36
58
Virtual Environment and .env File
08:44
59
Configuring the Page
10:15
60
Session State and Vector Store
08:07
61
Start Building the App: Sidebar
05:44
62
Building the App: Chat Inputs
05:07
63
Building the App: Bot Common Kwargs
10:00
64
Building the App: Bot Answers
09:37
65
Building the App: System Instructions
05:58
66
GitHub Repository
06:27
67
Deploying to Streamlit
03:03
68
Overview: Working With Unstructured Data
03:37
69
Introduction to Langchain Library
07:27
70
Excel Data: Best Practices for Data Handling
06:42
71
Initial Setup for Data Processing
10:47
72
Loading Data
06:37
73
Developing a Retrieval System for Unstructured Data
06:56
74
Building a Generation System for Dynamic Content
03:39
75
Building Retrieval and Generation Functions
09:05
76
Working with Word Documents
04:55
77
Setting Up Word Documents for RAG
11:49
78
Working with PowerPoint Presentations
04:45
79
PowerPoint Setup for RAG
05:33
80
Working with EPUB Files
04:59
81
EPUB Setup for RAG
04:16
82
Working with PDF Files
04:22
83
PDF Setup for RAG
09:56
84
What Did You Learn in This Section?
03:57
85
Exercise: Imposter Syndrome
02:57
86
Overview: Multimodal RAG
03:39
87
Introduction to Multimodal RAG
05:59
88
Setup and Video Processing
05:24
89
Extracting Audio from Video
08:45
90
Compressing Audio Files
04:18
91
Transcribing Audio with OpenAI Whisper
10:08
92
Whisper Model
06:32
93
Extracting Frames from Video
05:50
94
Introduction to Contrastive Learning
05:15
95
Understanding the CLIP Model
05:23
96
Tokenizing Text for Multimodal Tasks
08:14
97
Chunking and Embedding Text
11:37
98
Embedding Images for Multimodal Analysis
08:37
99
Understanding Cosine Similarity in Multimodal Contexts
06:47
100
Applying Contrastive Learning and Cosine Similarity
10:27
101
Visualizing Text and Image Embeddings
11:12
102
Query Embedding Techniques
04:13
103
Calculating Cosine Similarity for Query and Text
11:48
104
GenAI Model Setup for Multimodal Tasks
04:56
105
Building a GenAI Model
07:12
106
What Did You Learn in This Section?
02:13
107
Project Briefing: Starbucks Financial Data
05:28
108
Transcribing Audio with OpenAI Whisper
11:23
109
Embedding Transcription with CLIP
07:36
110
Converting PDF to Images
05:58
111
Embedding Images for Multimodal Analysis
04:59
112
Retrieval System
17:14
113
Preparing Context
05:00
114
Generative System
12:47
115
Game Plan for Knowledge Graphs with LightRAG
02:20
116
Knowledge Graphs
07:20
117
Knowledge Graphs vs Embeddings
08:50
118
LightRAG Setup
05:56
119
What is LightRAG?
04:41
120
Setting the Working Directory
05:50
121
Local RAG
08:41
122
Knowledge Graph Visualization
12:17
123
Global and Hybrid RAG
07:13
124
Naive, Mix and Bypass RAG
03:36
125
Overview: Agentic RAG
02:52
126
AI Agents
07:52
127
Agentic RAG
05:45
128
Setup, Data Loading and AgentState
06:50
129
State Management and Memory in Agentic Systems
07:55
130
Greeting The Customer
08:05
131
AI Agent that Checks the Question
07:04
132
AI Agent that Assesses the Validity of the Question
07:18
133
AI Agent that Generates the Answer
12:20
134
AI Agent that Improves the Answer
05:33
135
Asking User for More Questions
11:22
136
Testing and Improving Agentic RAG
05:42
137
Agentic RAG Recap - Key Learnings and Next Steps
06:18
138
Preparing the Prompt with ChatGPT or Gemini
18:15
139
Game Plan for Deploying Agentic RAG
01:04
140
UX Mock Ups with Stich
03:06
141
Setting Up with Cursor
02:59
142
Testing the App Locally
21:24
143
Final Debugging
04:13
144
Push to Github
04:42
145
Deploying to Vercel
01:49
146
Testing the App
04:11
147
Game Plan for RAGAS
01:54
148
Assessing RAG with RAGAS
06:14
149
RAGAS Setup
07:45
150
RAG
05:24
151
Synthetic Data
03:37
152
Generating Synthetic Data
07:03
153
Answering Synthetic Dataset
05:14
154
ROUGE (Recall-Oriented Understudy for Gisting Evaluation) Score
05:33
155
ROUGE
13:50
156
LLM-Based Assessment
06:08
157
Simple Criteria Score - Part 1
05:36
158
Simple Criteria Score - Part 2
05:40
159
Factual Correctness
05:17
160
Rubrics Score
04:53
161
Semantic Similarity
04:47
162
Factual Correctness
04:58
163
Context Precision
03:13
164
Semantic Similarity
03:12
165
Context Recall
06:22
166
Context Precision
05:58
167
Response Relevancy
04:37
168
Context Recall
04:56
169
Response Relevancy
06:23
170
Key Learnings and Outcomes: RAGAS
03:18
171
Thank You!
01:18
Unlock unlimited learning

Get instant access to all 170 lessons in this course, plus thousands of other premium courses. One subscription, unlimited knowledge.

Learn more about subscription

Related courses

Frequently asked questions

What prerequisites are needed before taking this course?
Before enrolling in this course, students should have a basic understanding of Python programming, as it is used extensively in the lessons. Familiarity with basic concepts of machine learning and AI models will also be beneficial, particularly when covering topics like transformers and attention mechanisms. Some lessons address API usage, so prior experience with APIs would be advantageous but not strictly necessary.
What type of projects will I build during the course?
Throughout the course, students will engage in practical projects including building a LinkedIn Post Writer App and a Mini Rubber Ducky AI. These projects incorporate the principles of Retrieval Augmented Generation (RAG) by integrating real-time data retrieval with generative model outputs. Students also learn to design user interfaces and deploy applications using tools like the OpenAI API.
Who is the target audience for this course?
The course is designed for aspiring AI engineers and developers who want to create smarter AI applications using Retrieval Augmented Generation. It is suitable for those interested in building applications like chatbots and search systems that require real-time data integration. The course also caters to individuals looking to enhance their skills in using the OpenAI API for practical applications.
How does this course compare in scope to other AI courses?
This course focuses specifically on RAG for LLMs, blending retrieval techniques with generative models to enhance AI application performance. Unlike more generalized AI courses, it covers niche topics such as multimodal RAG and working with unstructured data formats. It goes beyond basic AI concepts to offer hands-on experience with the OpenAI API and building end-to-end RAG systems.
What tools and platforms will I learn to use?
The course extensively covers the OpenAI API, teaching students how to set it up, write effective prompts, and configure model settings. Other tools include vector stores for efficient data retrieval and various programming exercises to integrate data from formats like PDF, Word, and Excel. Students also learn to use image processing techniques within the RAG framework.
What topics are not covered in this course?
The course does not cover deep learning model architecture design from scratch or advanced mathematical foundations of machine learning algorithms. It focuses on practical applications of RAG and does not delve into the development of custom AI models or training large datasets using frameworks like TensorFlow or PyTorch.
How much time should I expect to commit to this course?
The course comprises 171 lessons, each designed to progressively build on the previous one. While the exact runtime is not specified, students should be prepared to commit several hours each week to fully engage with the material, complete practical exercises, and participate in project work. The time commitment will vary based on individual learning pace and prior experience.