Inspiration / What it does

People often ask, what's the best financial option FOR ME? Our end product goal is to be TD based customer's go-to mobile application to receive financial advice based on their spending habits. Despite the fact that banking services are widely used by the majority of population, not everyone is aware of the existing options, and what each option is most suitable for them.

After being introduced to TD's Davinci API, we aimed to streamline the process of informing the customers of better financial choices by using select machine learning algorithms to analyze the TD's big data on customers' data and their transaction histories.

How we built it

The product is divided into a front-end componenet (Android Mobile Application (Java 8), Google Maps API) and a back-end component (Node.js, Express.js, Mongo-DB) that uses data (TD-Davinci API) analyzed using unsupervised learning (Scikit-learn).

Challenges we ran into

  • Figuring out suitable machine learning algorithm to find relationships among the datasets.
  • Building modular and scalable mobile application to provide positive user experience.
  • Pre-processing and analyzing big-data of TD's customer information before training machine learning model.

Accomplishments that we're proud of

  • Learned Android mobile application frameworks and features and created easy-to-use mobile application.
  • Being able to manage multiple subsystems
  • Using a structured software stack to build working front/back-end points
  • Analyzing customer data using an unsupervised learning algorithm and applying knowledge gained from the analysis in real application
  • Designed a simplistic UI that is intuitive and easy to interact.

What we learned

  • Applying unsupervised learning into real life cases, and visualizing/analyzing data
  • Features that are provided by the various APIs and how to implement them

What's next for For Me

  • Learn more in depth about different machine learning algorithms, evaluation techniques, and analyzing/visualizing methods.
  • Research more APIs that are accessible to incorporate them into future events.
  • Building more scalable mobile application accessible by various platforms (Android, iOS, etc.)
Share this project: