CashCast

Forecast Your Finances, Stress Less.

Inspiration

As busy college students, we constantly juggle classes, clubs, work, and social lives. With so many to-dos, keeping track of personal finances often falls to the bottom of the list. We wanted to create a simple, automated tool that helps students stay financially organized without extra stress. By forecasting spending and accounting for recurring bills, our app reduces the mental load of budgeting so students can focus on what matters most.

What it does

Our web app, CashCast, allows users to input their account number and instantly generates a personalized financial dashboard. Using real transaction data, CashCast:

Tracks income, purchases, and recurring expenses like rent, utilities, and subscriptions.
Automatically subtracts monthly bills from available funds.
Forecasts spending patterns based on recent history to predict when a user may leave the “green zone” of financial security.
Visualizes data in clear, easy-to-read charts that empower students to make better financial decisions.

The goal is to provide peace of mind by turning raw account data into actionable insights.

How we built it

Our team collaborated using Google Colab to quickly prototype and test ideas in Python. We connected to Capital One’s Nessie API to pull mock customer and account data, which gave us realistic transactions to work with. We then:

Designed wrappers around the API to fetch balances, bills, deposits, and purchases.
Simulated paychecks and recurring expenses to mimic a student’s financial routine.
Implemented forecasting logic using Python libraries such as pandas and numpy to analyze transaction history and project future balances.
Built an interactive dashboard with Streamlit, which made it easy to display forecasts, graphs, and alerts in real time.

Challenges we ran into

As first-time hackathon participants with limited coding experience, we faced multiple hurdles:

Technical setup: Getting comfortable with APIs, setting up keys, and understanding how to fetch real data. -Integrating Python and Streamlit: The most time-consuming portion was getting our python to run with our changes in Streamlit.
Debugging: Handling errors like 404s when working with mock IDs and learning how to navigate API documentation.
Time pressure: Balancing learning brand-new tools (Python, Streamlit, Colab) with actually building a working demo in under 72 hours.

Despite these challenges, persistence and teamwork helped us push through and deliver a functional product.

Accomplishments that we're proud of

We’re incredibly proud that, in just three days, we:

Successfully connected to the Capital One Nessie API and pulled realistic financial data.
Built forecasting logic that translates raw data into meaningful insights.
Deployed a working Streamlit dashboard where users can interact with our app.

Even more importantly, we grew as a team — going from idea to implementation while supporting each other’s learning curves.

What we learned

This hackathon taught us far more than just technical skills. We learned:

How to read and use API documentation to connect external data sources.
How to use Python libraries like pandas for data wrangling.
The importance of scoping — breaking down a big idea into smaller, achievable milestones.
How valuable collaboration is, especially when teammates are learning different skills at the same time.

What's next for CashCast

We’re excited to continue developing CashCast beyond the hackathon. Our next steps include:

Improving the frontend: Rebuilding the app using Flask and React to provide a smoother, more polished user interface.
User customization: Allowing users to categorize their own expenses, set savings goals, and receive notifications when they approach a threshold.
Mobile deployment: Making CashCast available as a mobile app so students can check their financial health on the go.