JellyQueue 🪼

JellyQueue is an intelligent meeting scheduler that uses the Gemini AI API to scan multiple team members' Google Calendars, resolve conflicts, check for weather conditions, and auto-generate the most optimal meeting time and Google Meet link, all wrapped in a conversational AI chatbot experience.

Whether you're working on a school project or organizing a club meeting, JellyQueue ensures everyone finds the perfect time slot, effortlessly.

How we built it

We used Figma for the wireframe and implemented our design using React for our frontend. The backend which handles the Gemini AI API and Google Calendar API were built as a Python Flask app. We connected the frontend and backend using the Flask-CORS package. Users login with their Google accounts via Google OAuth 2.0.

Challenges we ran into

AI-prompt engineering specifically for scheduling purposes Integrating multiple APIs that needed to communicate with each other

Accessing the Freebusy component of Google Calendar API: As we could not get proper returns from the Google Calendar API, we decided to write our own JSON data returns to test our AI chatbot when scheduling
Time slot conflict resolution requires access to the availability of multiple users
Creating an HTTPS server for Google OAuth 2.0 Unable to implement weather-aware scheduling due to time constraints

Accomplishments that we're proud of

Applying the knowledge of protocols and security standards learnt from coursework
Creating a cute UI and prioritize user experience
Creating a bridge between front-end and backend components

What we learned

Organizing the interactions between the many APIs with a diagram
Abstracting each interaction for easier debugging and more efficient teamwork
To be adaptable under time constraints
Navigating the Google Cloud console

What's next for JellyQueue

Implement weather-aware scheduling: using the National Weather Service API
Add NLP conflict parser: Users can type “I’ll be late Thursday” and it understands and updates automatically.
Add ML optimization: Use reinforcement learning to find optimal scheduling patterns (like minimizing rescheduling over time).