We planned to make a chat bot, using IBM Watson's revolutionary APIs (in particular Conversation), to give financial advice.

What it does

It give the location of a Capital One ATM or branch closest to a location specified by the user, within a range specified by the user, through a minimalist chat interface.

How we built it

We started with the Watson Conversation API that we used to train in a workspace to recognize simple, potential lines of dialogue flow between a user and the bot. Then, we incorporated Capital One API to retrieve the location of ATMs and branches within a radius of a location. Finally, we used Flask to generate a minimalist web interface between the chat bot and user.

Challenges we ran into

We ran into major challenges when attempting to move our app from the console to the web through Flask, which involved a restructuring of the python file that handled the interaction between IBM Watson's Conversation API and the user's messages. Also, there were some minor problems using the Capital One API, primarily due to the lack of info in the API (it seemed to only have locations of ATMs and branches on the East Coast).

Accomplishments that we're proud of

Given that this was our first hackathon project, we are extremely proud that we were able to generate a working product that will can be extended to cover more ground. Furthermore, we were able to incorporate multiple APIs into one app, which is definitely important in industry software development.

What we learned

We learned a variety of different technologies, ranging from using IBM Watson's API, brushing up on Python skills, and implementing a web app using Flask, HTML, and CSS. Furthermore, we realized how to work together as a pair, with one of us coding up the algorithm to determine the closest ATM and the other training the Watson Conversation workspace to be able to maintain a conversation with an user.

What's next for W-ATM-son

The next steps for W-ATM-son would be to add more functionality, such as using Nasdaq and Blackrock's Aladdin API to recommend stocks, and using IBM Watson's Alchemy API to analyze the content of financial news articles, and using Paypal API (through RapidAPI) to be able to make payments, among other related financial tasks and pieces of information.

Share this project: