AWS Unveils PartyRock for Building Generative AI Apps with Amazon Bedrock Integration
To improve your local-language experience, sometimes we employ an auto-translation plugin. Please note auto-translation may not be accurate, so read original article for precise information.
AWS launched its new generative AI app PartyRock, for anyone to learn about prompt-engineering in an intuitive and fun way.
Amazon Web Services (AWS) recently announced the launch of PartyRock — a platform designed for individuals interested in exploring generative AI while creating engaging applications. According to AWS, users would be able to enjoy the process of experimentation, learning prompt engineering and build mini-apps through the PartyRock.aws platform.
The company said users can achieve this without coding or setting up an AWS account. The platform will also allow users to start with pre-shared apps, providing a collaborative and creative environment.
“PartyRock has been designed in a way that is friendly and accessible to builders of all skill levels, particularly those who have limited experience with generative AI. Even if you have no coding experience, you can use text-based prompts to experiment with foundation models and create your own generative AI powered apps,” Mike Miller, Director of product management at AWS, told Metaverse Post. “Developers can explore model capabilities and refine prompt engineering techniques in PartyRock without managing API calls.”
To initiate the process, users can visit PartyRock.aws, log in using their Apple, Amazon or Google accounts, and access the PartyRock dashboard. From there, they can either review sample apps or choose to build their own by clicking on the “Build your own app” option.
Users have the option to describe the app they want to build, leveraging PartyRock’s generative AI for a quick start or they can opt for a manual, widget-by-widget creation. The platform aims to simplify generative AI application development process, making it accessible to a wider audience.
“PartyRock utilizes available foundational models from Amazon Bedrock to power applications. Users can selected any LLMs available in Amazon Bedrock, even like those from Cohere or AI21 Labs,” said AWS’s Miller.
Create Customized Generative AI Apps with PartyRock
AWS’s senior editor Jeff Barr recently explored PartyRock’s capabilities during the AWS re:Invent event. Detailed in a blog, Barr — amidst managing multiple blog posts, utilized PartyRock to generate a snarky response for impatient colleagues.
Barr’s app “Snarky Patient Blogger”, showcased two AI widgets available on the platform: user input and snarky response. The user input widget utilizes the Amazon Bedrock InvokeModel function, specifying the Claude v2 model and a prompt referencing the user input itself.
“PartyRock is an educational tool for providing users access to learn about prompt engineering through experimentation in a foundation model playground built on Amazon Bedrock. While it is not intended for in-production use cases, users can mix and match with different foundation models available today in Amazon Bedrock,” AWS’s Miller told Metaverse Post. “User can chain inputs into prompts and use generated text as input for other widgets by referencing them together using the “@” syntax.”
Barr also demonstrated how users can enhance created generative AI apps. He incorporated text generation and image generation widgets to visualize and represent the response. The resulting app was a smooth, user-friendly interface — demonstrating PartyRock’s potential for diverse applications.
PartyRock also comprises of additional development features such as creating an empty app, remixing existing apps, integrating chatbot widgets and utilizing advanced settings for widgets like text generation.
The “Remix” feature allows users to take someone else’s app, and then edit or tweak it to make it their own, while “Snapshot” feature to allow users to easily share the prompts and outputs with others from their created apps.
“Our intent was to create a frictionless experience, such that users don’t need to be a developer or ML engineer to start creating their own apps. They can just write a text-based prompt to describe what you want your app to do in the PartyRock app builder. After seeing how people were using this tech internally, we prioritized making it super easy to share apps and get inspired by others,” Miller added.
AWS announced it will offer a limited-time free trial for new PartyRock users, allowing them to learn fundamental skills without the need for a credit card or AWS account. Credit consumption will be based on input and output tokens, as well as generated images.
PartyRock usage is calculated via the input tokens, output tokens, and generated images for apps that you run. In a foundation model, a token is comprised of a few characters and refers to the basic unit that a model learns to understand user input and prompt to generate results.
“In PartyRock, input tokens are used based on the prompts you provide.This includes using the app builder to build apps, as well as adding text into an app widget to return a result. Outputs consist of the generated results that the foundation models working under the hood in PartyRock,” AWS’s Miller told Metaverse Post. “Credit usage varies by model to help users build intuition on cost consideration when using generative AI.”
Moreover, the company plans to introduce more models over time, expanding the platform’s capabilities. AWS said it is actively working on introducing new widgets and features to enhance PartyRock’s functionality.
“As more foundation models become available on Amazon Bedrock, we’re looking forward to adding these to PartyRock so that everyone can learn more prompt engineering through experimentation in a fun and frictionless way,” said Miller.
The generative AI platform signifies a step forward in democratizing AI application development, offering a user-friendly interface and diverse features for both beginners and experienced developers alike.
In line with the Trust Project guidelines, please note that the information provided on this page is not intended to be and should not be interpreted as legal, tax, investment, financial, or any other form of advice. It is important to only invest what you can afford to lose and to seek independent financial advice if you have any doubts. For further information, we suggest referring to the terms and conditions as well as the help and support pages provided by the issuer or advertiser. MetaversePost is committed to accurate, unbiased reporting, but market conditions are subject to change without notice.