Aeryth

Protector (Aegis) of Routines (Rythmos).

Comment

Aeryth - Aegis Rythmos (Protector of Rhythms)

Inspiration

What inspired me was something simple yet universal - Procrastination. From my friends, siblings, and even myself, I’ve seen how easily time slips away while we delay what matters most. When I first read the challenge prompt, I immediately knew I wanted to build something that helps people fill their empty time with purpose an anti-procrastination companion powered by AI.

What it does

As the name suggests, the main goal of Aeryth is to help users stop procrastination. But how can we actually achieve that?

The first step will be self-reflection, understanding our time wastage, recognizing our guilt, and visualizing what could have been achieved. That’s why I implemented explore topics with a personalized tone, digital sticky notes, calendar, and a personal diary system to help users stay aware of their goals and progress. These are supported by an AI summary that analyzes activities and provides insights for improvement.

However, as we could guess, self-discipline alone isn’t enough. If that were sufficient, procrastination would have been solved long ago. This realization led to the creation of Aeryth’s core feature - a nagging AI companion that sends personalized, adaptive notifications in different tones based on the user’s character.

Whether acting as a friend, coach, or even a parent, Aeryth evolves with the user’s responses, gently pushing them toward consistency and purpose. Over time, it transforms motivation from something external into something self-sustaining.

How I built it

Aeryth began as a web app designed to motivate users through personalized, AI-driven notifications. I integrated Gemini Nano’s Prompt API for generating context-aware reminders, combined it with Chrome extension scripts for background operations, and used Firebase to handle synchronized data across sessions.

Aeryth exists as both a web interface and a Chrome extension, bridging the gap between convenience and persistence, ensuring users receive timely motivation whether the app is open or not.

Challenges I ran into

One of the biggest hurdles was that web apps cannot natively send background/personalized notifications. To solve this, I experimented with service workers and later integrated a Chrome extension module that handles notifications while syncing with the main app. This required rethinking my data flow since I couldn’t rely on local storage anymore, so I shifted to Firebase for shared data persistence.

Each problem led to another lesson - from debugging API calls to optimizing extension manifests, and understanding how to safely connect frontend logic with browser permissions.

Accomplishments that I am proud of

I am proud that I was able to complete the prototype I had envisioned at the start of this challenge. Even though the model is not yet fully refined, it is already a usable and functional version of my concept.

What I learned

More importantly, through building Aeryth, I learned an incredible amount, starting from working with new AI frameworks to integrating Chrome extensions, handling real-time data, APIs, and undrstood that I have a very long way to go.

What's next for Aeryth

In the future, I plan to enhance Aeryth’s summary feature by integrating a dedicated Summary API for more accurate, context-aware reflections of user activity and Grammar API for some local corrections. I also aim to improve the notification system, making it more adaptive and natural in tone. Along with the available features I would also like to add a Pomodoro timer feature that not only tracks focus/break intervals but also uses your AI module to give smart, mood-based, or motivational feedback.

One of my broader goals is to see Gemini Nano evolve into a truly compact mobile framework. While 22GB is already impressive for an on-device AI model, it could be made even more accessible if users could select specific topic modules to download — essentially a mini-model of Gemini Nano tailored to personal needs and use them in diffrent devices.

If this approach becomes feasible, having a personalized reminder and reflection app powered by Aeryth on mobile devices with privacy would unlock its full potential — creating a companion that travels with the user and continually adapts to their rhythm.

Built With

Share this project:

Updates