Inspiration
It is incredibly easy to lose track of time. Many people find themselves wasting hours on their phones or on aimless pursuits—time that could be invested in bettering themselves, supporting their communities, or connecting with loved ones. The core inspiration for Time Bean came from a personal place: my project partner, Damian, wanted to spend more time visiting his grandmother, but continuously struggled to find the necessary structure in his day-to-day life to make it happen consistently. We realized that if we could build a tool to actively budget our free time based on what actually matters to us, we could solve this problem not just for Damian, but for anyone struggling with time management.
What it does
Time Bean is a smart scheduling application that uses an AI agent to help users budget their free time for the activities they truly value. Users start by creating a "Priorities" list, ranking how important each activity is to them. The app features a robust calendar where users can add their concrete, fixed events (like work, school, or appointments).
The magic happens when the user asks the app to auto-fill their schedule. The AI agent acts as a central brain, calculating the gaps in the user's day, analyzing the location data for both fixed events and priority items, and budgeting travel time. It then makes intelligent decisions to suggest priorities that fit perfectly into those time gaps, maximizing the user's ranked preferences while ensuring the logistics are geographically sound.
How we built it
We developed Time Bean using a 100% TypeScript and JavaScript stack. The frontend was built using the React Native framework with Expo to handle cross-platform mobile rendering. For our backend and database management, we utilized Firebase (specifically Firestore) to handle user authentication and store user events, locations, and priority rankings.
To bring the "smart" features to life, we integrated the Gemini API as the core scheduling agent, alongside spatial data concepts inspired by the Google Maps API to handle location tracking and travel logistics. Throughout the development process, we accelerated our workflow and problem-solving using AI developer tools, including GitHub Copilot in VSCode, GPT-4o, and Gemini.
Challenges we ran into
While we encountered our fair share of technical bugs, the most difficult aspect of this project was mastering collaborative team development. Working simultaneously in a shared repository taught us trial by fire about version control. We had to learn how to actively avoid and resolve Git merge conflicts, heavily comment our code, and maintain clear communication so that we could seamlessly understand and integrate each other's contributions without breaking the application.
Accomplishments that we're proud of
We are incredibly proud of the AI scheduling agent at the top of the program. Building an engine that successfully consolidates raw calendar data, location constraints, and user preferences into a clean, structured prompt—and then parses the AI's response back into a functional UI—was a massive technical hurdle. We are also proud of the app's design. Calendars are inherently limiting and rigid from a UI perspective, but we managed to create a consistent, creative, and highly responsive interface that feels modern and intuitive.
What we learned
This project was a masterclass in end-to-end software engineering. We vastly expanded our knowledge of full-stack development, asynchronous state management in React, and NoSQL database architecture using Firebase. Most importantly, we learned the critical fundamentals of professional development: how to build, debug, and ship software collaboratively in a shared repository.
What's next for Time Bean
Looking forward, we want to expand the capabilities of the Gemini agent. We plan to give it deeper access to historical scheduling data so it can proactively suggest a wider variety of activities to prevent user burnout. Additionally, we aim to improve the app's cross-compatibility to ensure a seamless experience across all devices and platforms.
Log in or sign up for Devpost to join the conversation.