#1 Generative AI On AWS-Getting Started With First Project- Problem Statement With Demo

Key Takeaways

hello all my name is krishak and welcome to my YouTube channel so guys I am quite excited to start the first project on generative AI on AWS Cloud uh in this specific video I will be talking about the problem statement that we will be solving not only that we will also go ahead and understand that what all ews services that we'll be specifically using to create this gen geni project and finally we I will also show you the entire end to endend demo uh which will probably show you that how this entire project Works um in the next video I will go ahead with the implementation so uh let me go ahead and share my screen over here and here you can actually see obviously I've defined the entire life cycle of the Gen AI project in my previous video again you can probably check out in the playlist itself and then over here uh this project we will be using all this Services you know and you may be thinking about f and all also that also we'll be discussing as we go ahead you know and all these things we will be specifically doing in the AWS now what is the project all about okay so uh this project what we will do is that uh we'll try to take up any use case Okay so there will be a specific use case over here that we will be solving so let's say this use case can be chat chat it can be a chat bot or it can be text summarization or it can be it can be code generation it can be different different use cases okay so let's say this is the use case that I'm trying to solve now to implement this specific use cases we will create a complete AWS ecosystem right a AWS ecosystem which will make sure that which will handle the complete endtoend project itself even you don't have to even go to your development environment you know so we will be taking taking care of multiple things over here now when I talk about AWS ecosystem these are the main services that we will be specifically using Okay so let's go ahead and start with the first time okay so let's say this is the client end okay obviously client end the client end basically means the user end here we will probably raise an API now how do we raise an API so let's say that whatever use case that we are specifically creating with respect to chatbot text summarization and G code generation or code generation it will be in the form of apis now this API can be probably created uh by using an amazing service that is available in AWS which is called as API Gateway okay API Gateway I will make you understand how to probably create this particular apis and what is API Gateway and all so this is one of the services that we will be specifically using now once we go ahead and probably create the API with respect to the API Gateway okay now this will probably trigger a event now in order to trigger a event as soon as we call the API this will go ahead towards another function which is basically called as Lambda function now this Lambda function will be the very important very important service because this will be responsible in calling the the foundation model services that is available in AWS Bedrock okay and now if I talk about AWS right there are two important services that we can specifically use one is AWS Bedrock now if you don't know about AWS Bedrock what does AWS Bedrock do what does AWS Bedrock do so this provides you lot of different Foundation models Foundation models to probably implement your generative AI use cases and this all Foundation models will be available in this AWS Bedrock itself you can probably create a client and you can call any of the foundation models that is available based on the region okay the second service that we specifically use is nothing but AWS Sage maker right now in AWS Sage maker also I can deploy my entire Foundation model okay I can deploy my entire Foundation model Now understand with respect to AWS Bedrock it is like pay as you go like you can use it as a service the number of based on the number of request you will be charged that much with respect to AWS sagea it provides you complete environment where you can also deploy your foundation models right AWS Bedrock directly gives you an API or it gives you an client end to probably invoke any of the foundation model with respect to all our playlist that we go ahead you know we will be first initially I will start with AWS bedrock and then I will also show you AWS Sage maker so from this Lambda function what I will do is that I will call my Foundation model from AWS Bedrock okay so from here we will be calling the all the foundation model we'll be invoking the foundation model right the other way is that directly host all the foundation model in AWS Sage maker so one way is through AWS Bedrock the other way is directly deploy a foundation model in AWS sagea and use that as an API and give any number of request that you specifically want right to access any kind of conversation or you really want to have any kind of conversation with the foundation model now with this AWS Bedrock we can further perform fine tuning and all also fine tuning and uh uh this use Cas I will probably show you in my next project right now in this specific project we will develop all this things right we will be creating an API Gateway we will be creating our Lambda function our Lambda function will be having the entire code right wherein we will be invoking the AWS Bedrock where where we will be invoking all the foundation models in the AWS bedrock and getting the response back and based on that response we will be giving it back to our API Gateway and finally we'll be able to give it back to the client as in the form of response so this is the entire architecture that briefly we will be developing and all these things will be developed in AWS itself as I said I'm just going to use vs code in just one of the steps other than that everything will basically be happening in the AWS platform okay and we can solve any of these use cases like chatbot text summarization and code generation uh my main aim is to use both open source models okay open source models and uh we can also use paid Source models so in uh Amazon Bedrock AWS Bedrock different different models are also provided to you like from anthropic you have Lama 2 Lama 3 all these kind of models you specifically have now uh these are the services as I said now let us go ahead and show you a demo right so this is what is the Lambda function that I have created this is my API Gateway and here you'll be able to see this is my entire code that I've actually written I will show you what this entire code is I will explain you each and everything as we go ahead okay uh along with that we will also be using some Cloud watch so that we can actually check out the logs and all so at the end of the day what we will be getting is that let's say I will be opening my Postman over here so with respect to this Postman This is My URL uh and understand one more thing guys when we do the deployment right of uh all this API Gateway when we do the deployment right we will be creating different different environments also so like Dev QA so that you also get that feel like how we specifically work in the prod okay so all this deployment techniques with respect to the API will also be shown okay now as I said uh this is one of the demo right so this is my Dev environment URL and I'm saying code generation and here I'm giving two keys like uh I want uh I want uh python code to implement binary search or let's say I want uh I want a python code to to sort to sort numbers in ascending order to sort numbers using selection s let's say okay so this is my code that I I basically want this is my prompt and the programming language is nothing but python so if I go ahead and click on send so here you'll be able to see that it will go ahead and hit my API so it will go ahead and hit this API Gateway then it'll go to the Lambda function then it will invoke the foundation model that is available in the AWS Bedrock it will give us back the code and it will then send it to the client end okay now here you can see code gener code generation is completed okay I have just given a message like this status is 200 okay now if you want to see the code uh what I will do and again you can also see CL Cloud watch here all the logs will get updated with respect to things so here you can see here sure here is the python code all the code is basically available over here so if I click over here all the code is also available and now you may be thinking where is the code saved so here I am going to call my S3 bucket now inside this S3 bucket you'll be able to see that I have this Bedrock course one the code output and here you can see this is the recent code right so the code is also getting saved in the S3 bucket so considering this uh as I said here we are also going to use S3 buckets S3 S3 buckets we'll save all the codes that is generated or all the chat that is generated all all the text that is summarized in the S3 bucket itself right so all this things and here I will also be showing you how you can update packages in Lambda function and all uh packages and all uh see I've used Lambda function a lot in my one of my previous company like Panasonic also I've used over there so this looks looks like a complete amazing end to endend project where you'll be learning all these things so I hope you're quite excited uh so let's keep the target 2,000 likes so that I will be uploading the next project as soon as possible so yes this was it for my side I'll see you all in the next video have a great day thank you and all take care bye-bye