Inspiration
This is one of the best projects since it tackles not 1, but 2 SDGs. The first one is Goal 9 - Industry, Innovation, and Infrastructure. The second goal is Goal 16 - Peace, Justice, and Strong Institutions, my project falls under this category because after researching, you can see the a sub category of this goal is the Department of Economic and Social Affairs(According to the United Nations).
The inspiration behind WiseGuard stems from a genuine concern for the well-being of the older generation in Singapore. Recognizing the rising prevalence of scams targeting seniors, the team aimed to harness the power of AI to create an innovative solution that educates and empowers seniors in identifying and avoiding scams. What it does
WiseGuard is an AI-powered website designed to provide practical knowledge on scams to the older generation in Singapore. The platform generates realistic scam scenarios using OpenAI's LLM. Users are then presented with quizzes related to the generated scenarios, focusing on highlighting red flags specific to each type of scam. The website allows users to cycle through six different types of scams, continually generating new scenarios for an interactive and educational experience.
How we built it
WiseGuard was built using a combination of Python, OpenAI, and Langchain for effective utilization of OpenAI's Language Model (LLM). Prompt design and engineering techniques ensured refined responses based on user-selected scam types. The backend, powered by Flask, processed LLM responses. For the frontend, Django, HTML, CSS, and Javascript were employed to create a user-friendly interface. Django's secure and rapid development capabilities facilitated account creation and login using SQLite. Docker and Render were used for seamless deployment, allowing users to access the website from anywhere.
Challenges we ran into
Challenges in developing WiseGuard included integrating the LLM effectively, designing an intuitive user interface catering to an older audience, ensuring the scenarios generated were realistic and relatable, and addressing potential ethical considerations in presenting scam scenarios. One of the largest challenges we faced was ensuring the responses from the LLM were of a specific format, as different parts of the output were presented on different parts of the website.
Accomplishments that we're proud of
The team at WiseGuard takes pride in creating a practical and engaging platform that empowers seniors with the knowledge needed to identify and avoid scams. The accomplishment lies in successfully integrating AI technology with educational components, creating a user-friendly interface, and addressing the specific needs of the older generation in Singapore.
What we learned
The development of WiseGuard provided insights into the challenges and considerations associated with integrating AI-focused solutions with real-world problems. It also gave us valuable experience in identifying various steps necessary to successfully build a full-stack project.
What's next for WiseGuard
The future for WiseGuard involves ongoing refinement and expansion. The team may consider gathering user feedback to enhance the platform's usability and effectiveness. Additionally, we might explore using larger or more fine-tuned large language models, implementing new types of scams, and possibly consider localization for other regions facing similar challenges.
Log in or sign up for Devpost to join the conversation.