dashboard
ai chatbot

Prospera

Prospera provides investing suggestions based on users’ current funds and their investing goals. Our web app features a built-in Gemini chatbot named Penny, tailored specifically towards budgeting and finance advice.

What Inspired Us

When researching existing finance apps, we noticed that most of them were focused on either budgeting or investing, not both. For example, Rocket Money is a popular budgeting app that helps users keep track of their purchases and savings, but it does not tell you what you can do with the saved money. Our web app Prospera does both to ensure users make the most out of their money. We wanted to especially help college students improve their financial literacy and make use of the advice that chatbots like our integrated Gemini can provide. When just starting out with investing, many beginners get overwhelmed by the many options for different types of investments and how to maximize their savings for their specific goals and needs. Thus, we wanted to provide our users a more accessible and personalized form of support that will ease their concerns regarding how to best save and/or invest.

Sponsor Challenges:

Capital One - Financial Hack
(MLH) Google Cloud - Gemini API
(MLH) GoDaddy Registry (prosperaplan.us)

How We Built It / What We Learned

We learned how to use Spring Boot and Gemini API and integrate them together into Visual Studio Code. With our GitHub repository to collaborate, we went through the process of setting up the environment of our web app and adding more interactive elements along the way. For frontend, we used HTML, CSS, and JavaScript, with Figma to design. For the backend, we used Java, Spring, JavaScript, and Rest.

Challenges

In the beginning of our ideation process, we were unsure of what environment to start out with, mostly between IntelliJ or Visual Studio Code. Then, we had issues with setting up the environment. With the Gemini API, we considered how to implement the API key in the code, as it would need to be used for user interactions to enhance their experience. Luckily, by printing error calls, the issue was resolved by troubleshooting the Gemini model version. Another major setback was the difficulty in understanding the material we were using (such as Nessie API) in such a short amount of time, trying to access their elements and how to implement them in the web app.