JoyFlow - Calendar Optimisation

UI design inspiration

Project Description - CS Girlies Annual Hackathon - Technology For Wellness

What does it do? This project is best explained to be a calendar optimiser, targeting the wellness of its user. It takes exported calendar data in form of .ics or .csv format and optimises it based on the users desires to give a more rounded and complete day beyond the work or business commitments. What inspired it? 24 hours can seem like a long time until you find out the time is 12 midnight and you still have things to handle, areas you have unknowingly ignored. From the perspective of a student and also some insight from a career individual, this is common. There are quite a number of things that matter to us which is my motivation. To be able to schedule it into our day or week, just as we do our career and business. What technologies did you use? Did you use AI? If you used AI, what model? Did you use a free or paid tier? I selected technologies and tools I had prior experience with. With the limited time allocated for the project, I took control over the concept of the project, its requirements engineering, UI direction, architecture and database design. Building only commenced after tool selected, UI design, database map out/visualisation; to a large degree had been done.

Technology Frontend: React Build: Vite Backend: Supabase Database: Supabase PostgresSQL Deployment: Vercel Version Control: Git, GitHub

AI - Development Assistant

Model: Codex Version: 5.6 Sol Reasoning level: Low Tier: Paid Which track are you submitting to? (You may submit to one main track and one or more bonus tracks) I am submitting for the Wellness track and the Best Use of AI track

Best Use of AI Track In this project, I used AI for code generation but also as an underlying tool in pulling different entities and giving suggestions from them. It helped:

convert natural-language desires into schedulable activities
compare desires with fixed calendar commitments
explain in understandable terms scheduling compromises, The underlying AI detail used is: OpenAI gpt-5.4-nano.