Inspiration
- We were inspired by the need for a simplified approach to discovering credit card and account offers, especially for young adults who may be navigating this process for the first time. Understanding credit card options can be overwhelming, and we saw an opportunity to empower young adults with a tool that makes the process easier and more intuitive.
- Our product will be increasingly beneficial as the credit card industry moves toward more personalized offers. As these offers become more tailored to individuals, it will become harder to rely on generic searches. Our platform ensures that users receive the most relevant, customized recommendations, saving time and reducing the complexity of comparing options.
What it does
- NextGen uses natural language processing and web scraping to help users discover credit card and account offers more easily. When someone interacts with the chat interface, the system uses NLP to figure out what they’re asking for—like rewards cards, low-interest options, or other specific features. On the backend, web scraping pulls real-time data from financial websites, collecting details about different offers like rates, fees, and rewards.
- From there, a recommendation engine compares the user’s preferences with the available options. The platform then presents these results in a simple, easy-to-read format, breaking down key features that the user values, to help them make informed decisions without the usual hassle.
How we built it
- We used different libraries and frameworks to build both the frontend and backend components of the platform. For the backend, we relied on Python and integrated web scraping libraries such as BeautifulSoup and Selenium to extract data from banking websites effectively. The frontend was primarily built with HTML and CSS for layout and design, with JavaScript added to enhance interactivity and improve the user experience. To link the frontend with the backend, we utilized the Flask framework, which allowed for smooth communication between the two.
- We used Gemini 1.5 Flash to make the interaction between the system and the user smooth and straightforward. It took the data we pulled through web scraping, which we saved in text files, and presented it directly to the user. This way, Gemini handled the communication side of things while we focused on collecting accurate data. It kept everything simple and made the platform feel more interactive and user-friendly.
Challenges we ran into
- An issue we faced was ensuring our files were properly organized and located in the correct directories. Misplaced files and incorrect paths caused several roadblocks, with front-end components failing to load and backend API calls not working. To resolve this, we focused specifically on the front-end HTML, CSS, and JavaScript code and made sure to follow a modular format. We established a clear folder structure, defined file paths carefully, and made sure all files were synchronized across all team members' computers. This approach streamlined our process and helped us avoid recurring issues, ensuring smoother development and deployment.
Accomplishments that we're proud of
- We are enthused with how we ended up linking the frontend to the backend because none of our team members have worked with building an application for this complex. It’s a very good learning process since it taught us a lot about how the functionality works within a website.
What we learned
- Throughout the project, we gained hands-on experience connecting the frontend and backend, learning how HTML, CSS, JavaScript, and Python all work together. We also learned how to use web scraping tools like BeautifulSoup and Selenium to pull real-time data from websites and store it efficiently.
- Integrating Gemini 1.5 Flash helped us understand how natural language processing works and how to make the user experience more interactive. We also ran into a lot of errors along the way, which taught us how to troubleshoot better—whether it was fixing file paths or debugging code with print statements and error logs.
What's next for NextGen Money
- Looking ahead, we’ve got some exciting updates planned. First, we’re looking to integrate our own machine learning (ML) model to make recommendations even more personalized. By analyzing user behavior and financial patterns, we’ll be able to offer smarter, more tailored suggestions that get better over time.
- We also plan to expand to other types of accounts, like savings accounts and CDs, so users have more options to explore beyond credit cards. This will make the platform even more useful for a variety of financial goals.
- Lastly, we want to stay on top of modern business trends to keep the platform fresh and relevant. Whether it’s adopting new tech or keeping up with the latest in the financial world, we’re committed to evolving with the times to offer users the best possible experience.
Log in or sign up for Devpost to join the conversation.