Most users do not get the optimal use of their banking account plans and they are unaware of better choices in the market. If a user does not need more than 35 transactions per month, s/he will save up to $200.00 annually without sacrificing anything. We hereby design BankingCat which analyzes user behaviour from banking statements, provide interactive conversation from ChatBot and outputs the best financial solutions across all Canadian banks.
How we built it
We use bootstrap and jQuery to beauty our frontend for landing page and chatbot window, and deploy the chatbot with natural language processing on Google's Dialogflow and connect to our backend Node.js server, which we host it on Google Cloud Platform with MongoDB.
Challenges we ran into
The problem occurs in data analysis. We have a difficult time trying to figure out the logic of analyzing banking statement.
Accomplishments that we're proud of
Our MongoDB and backend web server are successfully deployed on Google Cloud Platform. We are able to send messages to the frontend. This excites us since this is the first time we manage to host backend on Google Cloud Platform.
What we learned
What's next for BankingCat
We will focus on brining advanced intelligence into BankingCat's data analysis logics by introducing Azure AI and cognitive services.