DeProductify

Ever feel too locked in? Want to be more performative? DeProductify is for you! Our app tracks how productive you are - once you're determined too productive, you'll enter the performative focus mode!

Comment

Inspiration

In a world were too many people are locked in, we saw a need for individuals to become more performative while taking a break. This led us to create DeProductify, the app that promotes unproductive work, and punished your productiveness by putting you into our performative focus mode. Our app lets you take a break in a calm and relaxing way, while listening to some of your favourite performative artists, or drinking matcha.

What it does

DeProductify is an AI-powered anti-productivity app that actively monitors your screen and system behaviour. DeProductify analyzes a host of different productive attributes that determine how productive you are, such as:

  • Visual/OCR triggers: If you're on Google Docs (bright screen), the amount of text on screen, keywords (math, lecture, academic)
  • Productive Applications: What apps or brower tabs you have open, how many tabs
  • Typing/Tab Overload: How fast you're typing or how consistently you're typing

Each of these stats track score. Once you reach a productivity score of 0.5, you'll enter the Performative Protocol. This causes a bunch of performative items to float around on your desktop, such as:

  • All tabs and applications to close on your computer
  • Opens a window with performative items, such as vinyls, wired earbuds, matcha, etc.
  • Plays a random song from Laufey, Beabadobee, Clairo

The only way to exit the Performative Protocol is to drink the floating matcha 3 times (click on it).

How we built it

Languages & Core Tech:

  • Python - Entire application stack
  • CustomTkinter - GUI design to enable/disable the tracker
  • Google Gemini AI - Semantic classification of screen content with local caching Detection & Monitoring:
  • PyTesseract - OCR text extraction from screenshots
  • mss & OpenCV - Screen capture and frame analysis
  • pygetwindow & pyautogui - Active window tracking
  • sounddevice - Microphone input monitoring Overlay & Effects:
  • Tkinter + PIL/Pillow - Transparent fullscreen overlay with rotating, bouncing images
  • pygame - Cross-platform audio playback
  • pydub - Audio processing (2x speed, volume boost for shake SFX)
  • psutil - Process enumeration and termination

Challenges we ran into

  • Closing all apps except the ones essential to the OS
  • Playing music
  • Having a score threshold that wasn't too high or low
  • Having the Performative Protocol window fill the screen
  • Getting Gemini to reliably detect words without using too many resources

Accomplishments that we're proud of

  • Able to track ~14 different triggers to determine how performative the user is
  • Once a score of 0.5 has been reached, Performative Protocol opens, playing a random song with floating performative icons
  • Gemini AI used to detect productive words on the screen
  • Works with macOS and Windows
  • Forces users to take a break as it closes every app

What we learned

  • Integrating Gemini with OCR to read keywords from the screen
  • Using OpenCV to detect brightness levels of the screen
  • How to terminate non-essential programs
  • Playing audio from mp3 files
  • Detecting keystrokes, what apps the user has open and tabs

What's next for DeProductify

  • Compete against friends -> share you score with your friends and determine who can get the most work done without setting off the Performative Protocol
  • Achievement system for taking break

Built With

Share this project:

Updates