As students in high school, we realize the importance of finance in the coming years of our lives. However, right now, investing is stocks is too risky for students like us, and we wanted to engage in the stock market while maintaining financial stability. The market size for financial services is large, yet, applications simply do not provide a reliable method for investing stocks. Previous services require expensive brokers to make decisions, and mutual funds are inflexible for the customer. Because of these pressing needs, we created PodStock.
What it does
PodStock allows users to invest in stock in "pods", or groups of friends. Using Machine learning and the Capital One API, PodStock recommends stocks based on the current market, and provides visual insights to the user. The game-changing factor for PodStock is the fact that financial risk is eliminated, due to the incorporation of a voting system within a pod. A user in a pod can propose to buy a stock, and the group can accept or decline this.
How we built it
We leveraged several APIs to provide the best experience to the user. By implementing HTML/CSS and a Node. Js server, as well as the Capital One API, Tensor Flow, and MongoDB for storage and authentication. We also used the Google Finance and Yahoo Finance API. MongoDB
- Used for authentication of users and login
- Stores information about “Pods”, including members and account value BigParser API
- Used CSV files given from Google Finance to parse through information regarding stocks
- Implementing BigParser API to parse for disparities in stocks, and to detect drastic rises and falls in stock price
- Using this, we provide recommendations, as seen in the Recommendations Tab Capital One API
- Through the Capital One Hackathon Api known as “Nessie”, we created multiple customers with checking accounts to simulate the withdrawal of cash from their accounts through Podstock
- When profits or losses are generated by the stocks, the API is once again used to deposit cash into the checking accounts of these customers from Podstock
Challenges we ran into
We had difficulty connecting the data with a user profile, so we were unable to create a settings tab. We also ran into challenges while using MongoDB, because our group had experience with Firebase.io.
Accomplishments that we're proud of
Before the hackathon, our team didn't use Github for an enterprise-scale application, but today we were able to leverage Git to write code effectively. In addition, we are proud of the MongoDB work that was done after the challenges we had in retrieving data from the database.
What we learned
We learned how to use API endpoints, and even apply our Artificial Intelligence skills to come up with machine learning and predictive analysis. In addition, we also learned a lot about finance, because we did extensive market research before coming up with our product.
What's next for PodStock
For PodStock, we want to actually allow users to input their credit card information through PayPal. We would also want to make more accurate recommendations, using different APIs.