Ducky

Inspiration

I remember awhile back I was listening to a freakanomics podcast that was on the Eliza project. If you are unfamiliar with the project, it was when a computer scientist in the 60's wrote a simple chatbot that acted as a therapist. While its responses were limited it had a unprecedented and unpredictable acceptance by the public. The scientist who wrote it found that people would pour their hearts out to this machine that barely could process what the user was saying.

The point of this all is that humans crave interaction and sometimes machines are a better candidate for us than another human. We can all admit its hard to listen to others peoples problems but sometimes people just want to be heard or listened to. That is where Ducky comes in. He is named after the famous rubber duck that computer scientists talk to to help debug their code. Well Ducky the chatbot is there to help people debug their problems and ultimately just be a friend for people to lean on. Ducky will never get tired of talking to you.

What it does

Ducky is a chatbot that uses NLP to understand the user and form responses so that it can accurately measure what the user is talking about. Ducky helps flesh out ideas and tries to approach your problems from all angles. Instead of giving you a solution Ducky simply helps you methodically think about your idea and form conscience and organized thought about it. Acts like these have been proven to really help people reach that cathartic release. Ducky is here to help guide people to their own epiphanies and conclusions.

How we built it

We built it using Node JS and an NLP library. We also used the Botkit framework so it can communicate in real time with Facebook messenger and Slack

Challenges we ran into

Getting the appropriate responses based on user input organizing the learning process for the bot having enough time to develop all the angles..we struggled to finally move forward with an idea Having the bot reach an end goal have the bot not repeat the same thing