Doomuguard

DoomGuard is an AI-powered Android application that detects unhealthy scrolling behavior and provides personalized digital wellness insights using the Google Gemini API.

Comment

Inspiration

Modern social media platforms are designed to maximize engagement, often trapping users in endless scrolling loops without them realizing the psychological impact. We noticed how people spend hours doomscrolling late at night, losing focus, productivity, and healthy sleep habits.

Most existing screen-time apps only show statistics — they don’t truly understand behavior or provide meaningful guidance.

We wanted to build something smarter.

That inspired us to create DoomGuard, an AI-powered digital wellness companion that not only tracks unhealthy scrolling habits but also uses Google Gemini AI to generate personalized behavioral insights and wellness recommendations.

Our goal was simple:

Help users reclaim their attention in the age of endless scrolling.

What it does

DoomGuard is an AI-powered Android application that detects excessive scrolling behavior and provides intelligent digital wellness insights.

The app:

  • Tracks social media usage patterns
  • Detects prolonged scrolling sessions
  • Identifies late-night doomscrolling behavior
  • Calculates a personalized Doomscrolling Score
  • Generates AI-powered wellness recommendations using Google Gemini
  • Sends smart intervention warnings during unhealthy usage
  • Includes a Focus Mode to help users reduce distractions

Unlike traditional screen-time apps, DoomGuard focuses on behavioral understanding rather than just raw statistics.

Example AI insights include:

  • productivity impact analysis
  • sleep quality warnings
  • focus recommendations
  • personalized wellness coaching ## How we built it We developed DoomGuard as an Android application using modern Android development practices.

Technologies Used

  • Kotlin
  • Jetpack Compose
  • MVVM Architecture
  • Google Gemini API
  • UsageStatsManager
  • Retrofit
  • Kotlin Coroutines

Development Process

First, we built a usage tracking system capable of monitoring app activity and detecting potentially addictive scrolling patterns.

We then designed a behavioral scoring system called the Doomscrolling Score, which evaluates:

  • continuous usage duration
  • nighttime scrolling
  • social media exposure
  • session frequency

The most important part was integrating the Google Gemini API.

We engineered prompts that transform raw usage statistics into meaningful AI-generated wellness insights. Gemini acts as a personalized digital wellness coach inside the app.

Finally, we created a futuristic cyber-wellness interface inspired by AI dashboards and cybersecurity systems to make the experience visually immersive and engaging.

Challenges we ran into

One of the biggest challenges was balancing technical complexity with hackathon time constraints.

Usage Tracking Complexity

Accurately detecting scrolling behavior on Android is challenging due to platform limitations and permission requirements. We had to simplify certain tracking mechanisms while still maintaining meaningful behavioral analysis.

AI Prompt Engineering

Another challenge was designing prompts that generated helpful, realistic, and personalized wellness insights rather than generic chatbot responses.

UI/UX Consistency

Creating a premium futuristic UI while maintaining performance and usability required multiple design iterations.

Time Constraints

Building a polished AI-powered Android application within hackathon time limits required rapid decision-making and careful prioritization of features.

Accomplishments that we're proud of

We are proud that DoomGuard evolved from a simple idea into a fully functional AI-powered wellness platform within a short time.

Key accomplishments include:

  • Successfully integrating Google Gemini into a real-world behavioral analysis use case
  • Creating a unique AI-driven digital wellness experience
  • Building a futuristic polished UI/UX
  • Designing a personalized AI wellness coaching system
  • Developing a working Android prototype capable of live demonstrations

Most importantly, we are proud that DoomGuard addresses a problem millions of people experience daily but rarely think deeply about.

What we learned

Through building DoomGuard, we learned:

AI Can Be Used for Positive Behavioral Change

Generative AI is not limited to chatbots — it can become a personalized wellness companion capable of meaningful behavioral guidance.

Prompt Engineering Matters

The quality of AI responses depends heavily on context and prompt design. Small changes in prompts significantly improved the usefulness of generated insights.

Simplicity Wins in Hackathons

Focusing on a clear problem and delivering a polished core experience was far more valuable than trying to build too many features.

UI/UX Has Huge Impact

A visually immersive interface dramatically improves user engagement and presentation quality during demos.

What's next for Doomuguard

We believe DoomGuard has strong potential beyond the hackathon.

Future plans include:

  • Real-time AI behavioral coaching
  • Emotion-aware wellness analysis
  • AI voice assistant integration
  • Cross-platform support
  • Wearable device integration
  • Productivity habit tracking
  • Community wellness challenges
  • Personalized focus routines
  • School and parental wellness systems

We envision DoomGuard becoming a next-generation AI wellness platform that helps people build healthier relationships with technology.

DoomGuard — Fight Doomscrolling. Reclaim Your Focus.

Built With

  • api
  • architecture
  • compose
  • gemini
  • google
  • jetpack
  • kotlin
  • mvvm
  • retrofit
  • usagestatsmanager
Share this project:

Updates