Introduction

0:00 He starts by explaining text embeddings, including Amazon Titan text embeddings, and he shows how to seamlessly integrate Langchain with Amazon Bedrock. You'll also learn about the world of multimodal large language models and discover the power of Amazon Nova for both text and image understanding. And the course also covers retrieval augmented generation, RAG, and advanced multimodal rag techniques. Equipping you with the skills to build intelligent contextaware systems. You'll get practical experience by developing an end to-end application that leverages Amazon bedrock agents and knowledge bases to automate and optimize insurance claim processes.

0:49 So get ready to improve your AI skills for impactful enterprise ready solutions. And thanks to Amazon for providing a grant to make this course possible. >> Thank you so much for joining this course. My name is Suman Dvnad. I am a principal developer advocate for AI and ML at AWS and I am excited to have you in this course. In this course, we will start our learning by diving deep into embeddings, one of the most important concept in the field of natural language processing and of course LLMs.

1:36 We will explore few of the algorithms like bite pair encoding or BPE and shall learn how to create embeddings using transformers and hugging phase. Finally, we will also learn how to use few of the embedding models like Amazon Titan embeddings using Amazon Bedrock which is a fully managed service for building generative AI applications with different foundational models at scale on AWS. After that we will move on to multimodal models and shall learn how they are pre-trained how they are used and how you can use them with Amazon bread.

2:00 We will also learn about different models and architectures like contrastive language image pre-training or clip and bootstrap language image pre-training or blip. And finally we will dive into the latest family of multimodal models from AWS called Amazon Nova. After that we will shift our gears and learn about multimodal rack or retrieval augmented generation and we will build a system that can extract information from your data which contains not only text but also images and tables. And finally, in the last section of the course, we will build an end-to-end application to automate the insurance claim process workflow using agents and knowledge bases on Amazon Bedrock.

2:47 I hope you will enjoy this course and have a great learning experience. And during this course, if you have any questions, feel free to ping me on LinkedIn or put a comment below. I will be very much happy to help you out. This is my LinkedIn ID. So please connect with me and share your learning and before you get started I would strongly recommend to clone this course GitHub repository and keep it handy for your reference as you go along with this course.

3:35 Thank you once again for joining this course and with that I will see you in the next lecture. In this section of the course we are going to dive deep into embeddings for large language models. in general embeddings for natural language processing. I thought that it might be