Watson as was wrong about them spending to much
User it telling the truth about not spending a lot of money
User spent money but it wasn't enough to send a message
Anger and fear define that someone is trying to lie about their purchase
Inputting user information for Capital One
Taking in a balance and after a transaction has been made it sees if too much money has been spent and it shows you your account balance
For people have have no self control when it comes to buying things, we wanted to make an application that knows when you're spending too much money and motivates you to manage your money.
What it does
This application will connect with your Capital One account so that it knows what your balance is and when a transaction is made. Once a transaction has been made, a notification will pop up prompting you to open the application. Once opened, Watson will ask a series of questions to determine weather the purchase you made was a reasonable or if it was an impulse buy. If it was an impulse buy and it is over your spending budget for the month it will send a message to someone that you know will confront you about your spending habits.
How we built it
** Capital One ** Using the Capital One's API we found out how to pull user's bank information. With this we made a python code that would update your total spending and send that information to Watson Watson Using IBM's Watson API we developed a chatbot using bluemix and a tone analyzer using IBM's github repository. Using that information we made an android application. Twilio Made an account with Twilio that able to send SMS messages using their libraries.
Challenges we ran into
We were unfortunately unable to get all three of the aspects linked together on the application because we were using different languages.
Accomplishments that we're proud of
Being able to come up with an idea that we would all have fun with and getting this far with no prior knowledge of APIs.
What we learned
We learned how to work with API and how to generate a conversation capable of extraction information.
What's next for MoneyTalks
The next thing would be to combine all of the aspects of the project, make a nicer UI to make our application look more professional, and then upload the application to the market get feedback form actual users.