How to Build Generative AI Apps Easily with Amazon BedRock and PartyRock?

How to Build Generative AI Apps Easily with Amazon BedRock and PartyRock?

Amazon is the world’s leading cloud computing solution provider. Amazon has introduced some very groundbreaking services and platforms this year. Let’s us talk about its most innovative services, BedRock and PartyRock that help to build generative AI apps.  

What is BedRock?

Amazon Bedrock provides a fully managed service, offering access to leading foundational models from AI companies through an API. Additionally, it includes developer tools to help build and scale AI applications. 

Features of BedRock

With Amazon Bedrock, you can explore the following capabilities: 

  • Text Playground – A text generation application in the AWS Management Console. 
  • Image Playground – An image generation application in the console. 
  • Chat Playground – A conversation generation application in the console. 
  • Examples Library – Access a repository of example use cases for reference. 
  • Amazon Bedrock API – Delve into functionalities using the AWS CLI or leverage the API to interact with base models. 
  • Embeddings – Utilize the API to generate embeddings from the Titan text and image models.
  • Agents for Amazon Bedrock – Construct agents for orchestrating tasks and managing customer interactions. 
  • Knowledge base for Amazon Bedrock – Leverage data sources to assist agents in finding information for customers. 
  • Provisioned Throughput – Acquire throughput resources at discounted rates for running inference on models. 
  • Fine-tuning and Continued Pre-training – Tailor Amazon Bedrock base models to enhance performance and create a superior customer experience. 
  • Model Invocation Logging – Capture invocation logs, model input data, and model output data for all invocations within your AWS account using Amazon Bedrock. 
  • Model Versioning – Benefit from continuous updates and improvements in foundational models to elevate your applications’ capabilities, accuracy, and safety. 

What is PartyRock?

PartyRock is a platform developed by Amazon that allows users to build Generative AI  apps in a Playground powered by Amazon Bedrock. This platform provides a space for developers to experiment with various foundational models, understand text-based prompting, and learn how to effectively chain prompts. With PartyRock, the possibilities are endless. 

PartyRock provides a range of features and functionalities that empower developers to create innovative and powerful apps. Let’s delve deeper into some of the key features and explore how they can be utilized to create unique applications. 

Widget Functionality

The Playground in PartyRock consists of various widgets that serve different purposes. These widgets include: 

    • User Input: This widget allows users to input data or requirements into the app. It serves as the starting point for the app’s functionality. 
    • Static Text: The Static Text widget allows developers to display fixed text or instructions within the app. It can be used to provide context or guide users through the app’s interface. 
    • AI-powered Text Generation: PartyRock utilizes advanced AI models to generate text based on prompts provided by the user or the app itself. This feature enables developers to create dynamic and interactive conversational experiences within their apps. 
    • Image Manipulation: With PartyRock, you can incorporate image manipulation capabilities into your apps. This opens up possibilities for creating apps that can generate, modify, and manipulate images based on user inputs or predefined parameters. 
    • Chatbot Integration: By integrating chatbot capabilities into your app, you can create conversational experiences that mimic human-like interactions. PartyRock’s chatbot widget allows developers to build intelligent and engaging chatbot functionalities effortlessly. 

Prompt Engineering

Prompt engineering is a fundamental concept in PartyRock that enables developers to generate meaningful and contextually relevant outputs. By carefully crafting prompts and leveraging the power of AI models, developers can create apps that produce accurate and desired results. 

Dynamic Variables Chain

One of the key strengths of PartyRock is its ability to chain variables dynamically, allowing the output of one widget to be used as input for another. This flexibility enables developers to create complex and interconnected apps that seamlessly flow from one widget to another, enhancing the overall user experience. 

Let's explore how to develop Generative AI apps using PartyRock

To use PartyRock, go and visit the link: https://partyrock.aws/ and Sign In  

Generative AI apps using PartyRock8

After logging in, you can access the PartyRock and build your App to get started. 

Here, we have created one simple AI-based application, HelathAdvisor 360, using AWS PartyRock. 

This App provides AI-powered healthcare insights. We have two options for creating an app using PartyRock. We can either describe the App you want, and PartyRock’s Generative AI will help kickstart the process, or we can build it step by step, adding each feature like building with widgets.  

Let’s go through a simple way where we don’t have to add widgets step-by-step. 

With the help of a prompt, as seen in the image below, one must explain how they want the application to be received. 

The prompt is: “Develop an AI-based healthcare application that significantly benefits people by providing personalized and insightful assistance. The application should utilize advanced algorithms to analyze health data, offer tailored recommendations, and enhance overall well-being. Consider incorporating features such as real-time health monitoring, predictive diagnostics, and user-friendly interfaces to make the App accessible and helpful to a wide range of users.” 

Generative AI apps using PartyRock7

After describing the app, you desire using the prompt, click on ‘Generate App,’ and within seconds, your application will be ready just the way you want it. 

Generative AI apps using PartyRock6

The App HealthAdvisor360 is ready within a few seconds. Now, we will give inputs to see if the output has sufficient snark for our needs: 

The Input provided was. 

Generative AI apps using PartyRock5

The received output is shown in the image below. 

Generative AI apps using PartyRock4

Isn’t that amazing? Let’s break it down and understand how it operates.  

This basic widget includes a call to the Amazon Bedrock InvokeModel function. The widget defines the use of the Claude v2 model along with a straightforward prompt, referencing the User Input widget. 

Generative AI apps using PartyRock2

You can play around by tweaking either of these elements, saving the modification, and seeing the updated result in a second. As an illustration, use the different models of Claude Instant, Claude, Jurassic, etc., to see the other responses. 

Generative AI apps using PartyRock 1

After successfully creating an application using PartyRock, we can share it with others by clicking the option ‘Make Public.’ 

Congratulations! People can now use and enjoy your AI-based Health Advisor app, which has just been released. 

Currently, AWS is letting new PartyRock users try it out for free, and you don’t need a credit card. Go ahead and start building your Generative AI app with PartyRock! 

Take advantage of this new offering from AWS and quickly build your AI application.  

Happy Coding!  

Are you ready to unlock the full potential of AWS Managed Services? Contact us at Futran Solutions to explore how our Cloud Managed Services can benefit your business. Our team of skilled professionals is committed to delivering exceptional support, security, and innovation to help you thrive in the cloud. 

Share:

More Posts

Twitter