Building a RAG Solution from Scratch

Course details

In this course, Axel Sirota introduces Retrieval-Augmented Generation (RAG) as a powerful technique for enhancing the capabilities of Large Language Models (LLMs). Learn the foundational concepts and practical applications of RAG, focusing on creating chatbots and decision support systems across various domains. Using the MIMIC-III dataset to create a healthcare chatbot that can answer questions or suggest a diagnosis as an example, get hands-on experience in building RAG systems with TensorFlow, Keras, and HuggingFace. By the end of the course, you will be equipped to deploy RAG solutions that integrate robust retrieval mechanisms with generative models, applicable in fields like healthcare, legal, and customer service.

Learning objectives
Understand the principles and architecture of RAG systems.
Implement RAG solutions using TensorFlow, Keras, and HuggingFace.
Develop and deploy chatbots and decision support tools.
Explore various applications of RAG across different industries.
Analyze research developments and future trends in RAG.

Concepts

Welcome

01 - Introduction to RAG solution from scratch
02 - Getting the most out of this course
03 - Version check

1. Introduction to RAG Systems

04 - What is Retrieval-Augmented Generation (RAG)
05 - Key components of a RAG system
06 - Comparison with traditional LLM approaches

2. Creating the Vector Database

07 - Demo - Setting up the development environment
08 - Data preparation and preprocessing Techniques
09 - Creating the vector database
10 - Demo - Implementing the vector database with MIMIC-III
11 - Solution - Building and using the vector database

3. Developing RAG for Chatbots

12 - Designing a chatbot architecture with RAG
13 - Demo - Integrating the generation component
14 - Demo - Full integration and testing
15 - Optimization and evaluation techniques
16 - Solution - Deploying a functional chatbot

4. Implementing RAG for Decision Support Systems

17 - Designing decision support systems with RAG
18 - Demo - Building the information retrieval system
19 - Demo - Implementing decision support logic

5. Research, Ethics, and Future Directions

20 - Current research and innovations in RAG
21 - Future trends in RAG development

Conclusion

22 - Course summary and next steps