The digital equivalent of yelling into a pillow, leaving only an AI-generated trace of the yell's core. Built with React, Ruby on Rails, and just enough irony to make it modern.
Void is a minimalist web app designed for venting into the abyss. Users submit short text "vents" which are processed by an AI model that extracts the core subject matter. These subjects are visualized in a live word cloud, offering a glimpse into the collective chaos. It's therapeutic, anonymous, and very unnecessary.
Think of it as a screaming room with data viz.
- Frontend: React, Tailwind CSS, Wordcloud.js
- Backend: Ruby on Rails 7, Sidekiq (for background jobs), Transformers via PyCall
- Deployment: Configured for Heroku
- Submit free-form emotional content (aka posts)
- AI parses the content using a transformer pipeline to extract the subject
- Background job processing ensures smooth UI performance
- Real-time word cloud updates with extracted themes/ subject-matter of the post
- Responsive, minimal interface designed for maximum venting efficiency
- A user submits a vent through the frontend form.
- The post is stored via Rails API and enqueued for processing.
- A background job (using
ActiveJob) calls a local AI model viaPyCallto extract the post’s subject. - The extracted subject is saved back to the post.
- All subjects from submitted posts are visualized in a live-updating word cloud.
To run the app locally:
# Install dependencies
bundle install
yarn install
# Set up the database
rails db:create db:migrate
# Run the dev server
rails s
You’ll need Python with the necessary packages (e.g. transformers, torch) to support the subject-extraction job.
This project was built as a toy app, but it still showcases integration of AI, async job handling, and a modern reactive frontend.
🧙♂️ Powered by:
- React
- Ruby on Rails
- Tailwind CSS
- Transformers by HuggingFace
- React Wordcloud
May your feelings be loud and your stack traces silent