Inspiration
In today’s economy, it’s common and realistic for people to have financial concerns for dating. Most want to find a lifelong partner with stable income, low debt, and bank account numbers that match up, or even surpasses their own. Through our dating web app, users' concerns about finances can be addressed by having the option to put in their own financial information, get evaluated by a score and get recommendations of possible partners that match not only their preferences but also financial stability.
What it does
First, users create an account by inputting their phone number, then email and password. The website will then lead to a questionnaire which asks for goals in relationship, living circumstances, all of which will be stored into a database table and evaluated by AI to make the best match. Then the user will be asked for their financial situation, including options like income, savings, debt, and risk tolerance, which will be taken into account in the final compatibility evaluation. Finally, we integrated a multi-agent AI function using an openrouter API, so we have Gemini nano write an overall summary for the profile, with an optional choice to do recording which will also be taken into consideration for the matchmaking with Llama detecting tones to assume additional characteristics. Then Mixtral will provide users with a financial scoring to 100 that indicates their overall performance in maintaining low debt and high savings. After the profile is completed, the user will receive a list of recommendations of users that match their overall preference, with both the personality and financial preference taken into consideration. There’s also an event planning chatbox function by deepseek which helps the user find the date event that most closely matches their budget (data from NYC events) and argue for prices. The AI serves as a memory allocator which remembers the users info upon creating the account, and remembers it throughout the chatting function For the models we’re using deep seek and chatgpt for the therapy, AI generated summary, and financial evaluation.
How we built it
CSS for forntend, Java. for backend, save sign-in info from user to database, implement API for chatbot
Challenges we ran into
Learning to connect database with IDE, implementing API
Accomplishments that we're proud of
Building a fully functional website on our first hackathon.
What we learned
Implementing the API key and utilizing different models
What's next for Marry like Berry
A more detailed market research on deciding the threshold of money for people to join the events Seeking collaborations with restaurants, event lounges, and social clubs. Profile picture
Log in or sign up for Devpost to join the conversation.