AnxietyBot

Inspiration

In today's world, there have been signs of increasing levels of anxiety and no one knows quite why. However, one fact that we know for sure is that people tend to reach out to people less because they fear judgment from the people who they might reach out to. Therefore, we've decided to make a bot that is easily accessible and will not judge you for what you will be telling the bot.

What it does

The bot aims to listen to first listen to the key word of I am feeling ____. Then it evaluates how you might be feeling by sending a request to the IBM Watson Tone Analyser to evaluate how you might be really feeling by the words that you might be using. Depending on how high your level of sadness, anger, frustration or anxiety might be, it'll send you a comforting text message to your phone via Twilio and a GIF via GIPHY that might give a laugh.

How we built it

We built this by using AWS with Alexa and coded everything in Python. Due to our lack of resources of being able to setup an environment to download the helper libraries on python, we used requests to the request on the API's that used such as GIPHY API, Twilio API, and IBM Tone Analyser. In addition, we are encrypting the gifs by an RSA algorithm so that you won't be concerned about people knowing what kind of gifs you might be receiving depending how you are feeling.

Challenges we ran into

We ran into problems as mentioned before where we didn't have enough time to setup the AWS development on our local machines and had to hack everything in one large file. This caused a problem in both workflow and neatness.

Accomplishments that we're proud of

We are proud of how this actually works and how this can be extended into other potential projects. In addition, by training Alexa with more skills, it can be used to lead into a dynamic conversation if extended further.

What we learned

We learned that it was at first difficult to set a plan in motion with a limited amount of time.

Built With

alexa
amazon-web-services
encryption
giphy-api
python
twilio

Submitted to

Fall 2017 hackNY Student Hackathon
- Winner Best Use of Encryption [Sponsored by Datto]

Created by

I worked on the integration of Twilio API and the IBM Bluemix Tone Analyzer API. There was a lot of learning curve in trying to integrate an API within AWS using requests. It was a lot of fun experimenting with a lot of API's and especially thrilling to see everything come together.

Yong-L
I worked on the Giphy API which we used through Twilio to send relevant memes to users. The language we used for the APIs is python. It was my first over night hackathon, so the biggest challenge was continuously working for such a long time and also working along side others as team. This was the first time I've worked with Giphy API and Amazon Alexa skills, it was a lot of fun.

yuanzhou15
I worked on the Amazon Echo integration, which includes getting input from the user, sending it over for tone analysis, and sending back responses to the user based on tone analysis results. The primary tools I used were the AWS Lambda and Amazon Alexa skills configurator. All programming was done in python. I also managed the team's efforts, ensuring we achieved our goals within the limited time.

Subhan Altaf
I worked on the encryption aspect of the project. This included the encryption that was done on the Alexa echo and the decryption that was done through the android application I created. I also helped with part of the AWS lambda function as well as debugging the final code.

Mikhail Kreytser

Updates

Yong-L started this project — Oct 15, 2017 08:56 AM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.