Inspiration
I grabbed a coffee with a friend recently and we began discussing the importance of inclusivity in the tech industry, beyond just gender diversity, to be more inclusive for people with disabilities. For some people with disabilities they may struggle with navigating home websites and consuming its content. I was inspired to create an application utilizing Generative AI to make the web browsing experience inclusive and friendly for everyone and help web developers.
Web accessibility is an important element to consider when bringing a website to life. However, it is a design element often left for last. Designers often struggle with it or the development team does not have the know-how. Through this application, the designers and developers can run their website URLs and check what improvements can be made to make their web design as inclusive as possible. This helps showcase how a Generative AI application can become part of a development team's toolkit.
What it does
A friendly gen AI app that helps developers and UX/UI designers instantly check for ways to improve their websites and make it more accessible.
How we built it
Getting started with generative AI prompt design and merging it with an user first approach
Challenges we ran into
Crafting the correct prompt to provide enough information for the user and specific to the website information can be difficult. There is a fine line between providing too much information and it can become useless. Thus, overcoming what type of bite-sized information is valuable to different types of audiences. This required learning how to utilize temperature and Top P variables along with the prompt design.
Accomplishments that we're proud of
First Generative AI App design and selecting the best LLM model for it
What we learned
Prompt designing and tailoring it for the audience that will be interacting with the application. Along with selecting the best LLM model and tweaking of its parameters for the desired outcome.
What's next for AWS PartyRock Accessibility App
Obtain more feedback from the users to learn how to best tailor the feedback provided. Right now it is focused on providing programmer friendly feedback so the changes can be updated on the website with minimal hassle.
Built With
- amazon-web-services
- partyrock
Log in or sign up for Devpost to join the conversation.