Inspiration
In the Muslim community or any religious community, ethics are extremely valued and are followed by the T. As the Internet becomes more unregulated due to the freedoms of the land, it is hard to make sure that the people of our minority are not marginalized content-wise. We should not have to adapt to American values, but instead, maintain and grow our own. This led to the creation of SERAT.AI. It is a series/stack of AI and API programs used to regulate the web to make it Muslim-friendly. It comes in an assistant style, so it has a chat and multiple variations of interaction.
What it does
Built with the API of Chat-GPT, we dig deep into the technologies real capabilities by commanding it to filter and execute based on user input and website scraping.
The first feature consists of a live-action website filter. Simply, it filters through a website or app to look for key "taboo" words based on Chat-GPTs' prompt judgment prerequisites.
The second feature is a pop-up. Powered by alAdthan and Google extensions, it will give Google notifications based on when the athan is called.
The third feature, an entertainment filter, is a chatbox where you can provide a name of a movie, and chat-GPT will use the text Python scraped from the reviews database to guide the user on whether the movie is appropriate or not. The program is automated to give a human-like (Islamic scholar-like in this case) review of its correlation or contrast with Islamic values, a relevant ayah to the matter, and a url to further investigate the film's contents for further assessment.
The fourth and final feature consists of a chatbox programmed to answer as an Islamic scholar based on age, providing a custom experience for all ages of our ummah.
How we built it
We used: ChatGPT-API and Python for every feature, except the pop-ups, which were using the aladthan.API filter. We created a working prototype through Figma while using Canva for digital art.
Challenges we ran into
Debugging just never gets easy, but. once you find a good mentor, it energizes you to continue the work to become better and complete the challenge.
The biggest problem we had to solve was creating the functions to scrap the websites and feed it into chatGPT. It was a lot of inspecting and analysis.
Accomplishments that we're proud of
As a very detailed oriented crew, we loved the way we branded our product, from colors to bios, everything was well-thought of why it would be there.
We also are happy to announce that is our first project using AI and APIs, and we feel like we have secured a new trick under our belts.
What's next for serat.AI
Inshallah, we will try to:
- Provide an Auth.O log-in
- Different UI/UX experiences
- Community Feedback Forums
- Implementing alathdthan.api
- App Development
- Listening to the judges, like
Log in or sign up for Devpost to join the conversation.