ANDINI

Talk to ANDINI, discover YOURSELF

Comment

Inspiration

From COVID-19 social isolation to UF Students having their spring break being moved to the end of a short winter break, mental health has plummeted at the beginning of the decade. Some, even with insurance, cannot afford to get help working through and organizing their problems. Even then, some who can afford it would rather chose not to because of societal stigmas of needing to seek help. Andini is meant to offer a non-invasive way for people to work through their issues and reflect.

What it does

Andini is a guided journal that starts off by asking users how their day has been and then proceeds to ask questions of increasing specificity based on their responses. The questions are chosen based on natural language processing, finding keywords and their associated sentiments and continuing the conversation based on those results. The user's overall mood is evaluated and logged so that they can keep track of their mood over a period of time, accessible via a standard username and password login.

How we built it

The natural language processing is accomplished using Google Cloud's natural language client library. Using Google's NLP, we extract the keywords and sentiment from the user's responses and then decide what question to ask them next. On the front end we've written an iOS application using Swift. The backend is using Django for the framework and CockroachDB for user data.

Challenges we ran into

Initially we were going to host the entire back end on Google Cloud. We ran into issues with integrating our app, which was deployed on Standard App Engine, with a CockroachDB installation on the cloud. Ultimately we decided that everything would be run locally out of our apartment.

Accomplishments that we're proud of

Two out of four of our team members only started learning python just recently, despite this they were extremely prolific in writing code for the question generation/selection algorithm as well as extremely influential in the app's initial concept. This app is truly something that we have worked hard for and it feels especially fitting because this will be the last SwampHacks for 3/4 of us.

What we learned

Especially when competing in a Hackathon, being able to pinpoint what tools will work together well quickly as part of your development is incredibly important.

What's next for ANDINI

Moving everything into the cloud. Making it deployable on iOS. Increasing the breadth of the questions and question selection algorithm. Can always be made better.

Slack Contact

Nabeel Farooq (Borjes)- swamphacks-vii.slack.com/team/U01KRD2GZM4 AZ Vasquez - swamphacks-vii.slack.com/team/U01KYCRB4UB

Built With

Share this project:

Updates