Finance For Dummies

Inspiration

The inspiration behind this project came from our shared passion for AI Implementation and the desire to create a user-friendly application that simplifies financial management, which often times seems too daunting to approach. We wanted to develop a solution that combines convenience, insightful advice, and user interaction to empower individuals in making informed financial decisions without any fear.

What it does

At the core of our platform is a sophisticated financial advisor powered by customer data. This intelligent advisor provides tailored and personalized recommendations, guiding each customer towards CapitalOne's exclusive money management tools. By leveraging customer insights, we empower individuals to make informed financial decisions and optimize their financial well-being.

In addition, we understand the importance of convenient access to funds. To address this, we have developed a powerful tool that enables users to locate the nearest ATMs swiftly. With just a simple click, our program identifies the three closest ATMs based on the user's current location. But we didn't stop there - we took it a step further. With a seamless button press, users can effortlessly obtain precise directions to the selected ATMs, ensuring a hassle-free and efficient banking experience.

By combining advanced financial guidance with effortless access to funds, our platform delivers a comprehensive and user-friendly banking solution. We strive to empower our community with the tools they need to manage their finances effectively, making banking easier, more convenient, and tailored to individual needs.

How I built it

We used the openai github repo to start with a openai flask app. We then changed the html to a banking UI using chatgpt and css, and then we started retrieving data from the Nessie API. We passed the Nessie API variables (nickname, deposit, etc...) into the generate prompt function in openai to make the chat personalized on the users data. for the backend, we stored the enterprise data in a .json file to later access and augment. For finding atm's near the user, we made scripts and used costumer api from CapitalOne to covert the user's address and nearby atms to longitude and latitude coordinates that can be visualized with google maps.

Challenges I ran into

During the course of this project, we encountered various challenges that slowed down our progress. With multiple aspects to our website, we faced routing issues that persisted even after making code changes. As newcomers to hackathons, resolving these issues took longer as we hadn't encountered them before. One significant hurdle was configuring our site to run smoothly on Azure. The process of linking our GitHub repository to Azure and generating a functional site required extensive trial and error, which proved time-consuming. Another critical task was ensuring our AI provided accurate responses to user inquiries. Fine-tuning the model involved dedicating time to refining it with multiple sample prompts, ensuring it delivered appropriate answers. Furthermore, implementing a service to locate the nearest three ATMs posed challenges. Understanding and finding alternatives to the Google Maps API proved to be a complex undertaking. Lastly, developing a ranking system to determine the closest APIs based on street addresses presented another significant challenge. Despite these obstacles, we persevered and found solutions to overcome them, ultimately advancing the progress of our project.

Accomplishments that I'm proud of

As first-time participants in a hackathon, we were thrilled by the progress we made. Our end product turned out to be a remarkable and multi-functional website, seamlessly integrating multiple APIs. One notable achievement was our successful utilization of the Nessie API and OpenAI API, which greatly contributed to the website's functionality.

We also overcame the challenge of getting the site up and running through Azure, a process that is often known to be time-consuming. Despite the difficulties, we persisted and achieved success, further boosting our sense of accomplishment.

Overall, our experience in the hackathon left us immensely proud of what we accomplished—a well-designed website that effectively utilized various APIs. We learned a great deal from this journey and look forward to applying our newfound knowledge in future projects.

What I learned

Throughout the project, we embraced a continuous learning process. We started by getting comfortable with Git for efficient collaboration. Then, we dived into the backend, implementing the Nessie API and the OpenAI API. We fine-tuned the OpenAI API to ensure tailored user responses. Once that was in place, we began building the website using HTML and CSS. Later, we migrated the site to Microsoft Azure, exploring the wide range of functions and options available. The project taught us the importance of connecting all aspects of a website as full-stack developers.

What's next for capitolOneBankAzure

As the future of banking gets more and more advanced and complicated, we decided to democratize access to banking by utilizing financial data to provide tailored personal responses to suit their needs. As individuals get more ways to invest their money and allocate their resources in different ways, it is important to stay on top of your account information with your very own AI implementation that works 24/7 365 days of the year, free of charge. Our mission is to further enhance our AI implementation to provide even more accurate and personalized responses based on users' financial data. By leveraging machine learning and natural language processing techniques, we will strive to make the AI smarter and more insightful, offering valuable financial advice and recommendations, along with eventually offering all of the security measures that banking apps need.

Lessons Learned/Project Reflection

Throughout the development of this project, we gained valuable knowledge and skills. Firstly, we learned about integrating HTML and CSS to create an engaging and visually appealing frontend interface. Secondly, we deepened our understanding of Python3 and Flask, using them to build the robust backend that handles user authentication and financial data processing. Lastly, we honed our ability to integrate external APIs, specifically the GPT API, to provide personalized financial advice.

Project Development

To build this application, we adopted a combination of technologies and frameworks. The frontend was developed using HTML and CSS, allowing for a responsive and visually pleasing user interface. On the backend, we utilized Python3 along with the Flask framework to handle routing, user authentication, and database interactions. The financial information was securely stored and accessed through appropriate APIs, ensuring data privacy.

Challenges Faced

During the project development, I encountered several challenges. One of the major hurdles was integrating the GPT API into the chatbox feature to provide accurate and relevant financial advice based on the user's banking information. This involved understanding the API documentation, implementing data preprocessing techniques, and handling API responses effectively. Additionally, ensuring data security and user privacy throughout the application posed another significant challenge that required careful implementation and testing.

Overall, this project provided an enriching experience, allowing me to explore various technologies, enhance my development skills, and tackle complex challenges to deliver a user-friendly financial management web application.