SleepSync

Inspiration

SleepSync was inspired by two things: an Iowa State article highlighting the severe sleep struggles students face, and the way my dad—and Dr. Viru “Virus” Sahastrabuddhe from 3 Idiots—master the art of short, efficient afternoon naps. Watching them treat time as a disciplined resource made me wonder how students could better use all the hours outside their fixed routines. That spark grew into the idea of an intelligent planner that protects sleep and structures the rest of life around it.

What it does

SleepSync is an AI-powered weekly scheduling system that generates realistic, minute-accurate timetables. It ingests class schedules, Google Calendar events, Canvas assignments, and personal habits (like gym time, commute, or lunch), then uses a constraint solver to produce a balanced weekly plan. The app automatically protects your sleep hours, distributes study blocks, adjusts for exams or new assignments, and visualizes everything in a clean 7-day timetable grid.

How we built it

The backend uses Python + FastAPI, feeding all data into a custom pipeline that normalizes Excel, ICS, and scraped Canvas inputs. We transform every event into unified time blocks and model the full week as a set of integer variables using OR-Tools CP-SAT. The solver enforces constraints like non-overlap, daily limits, travel gaps, and protected sleep windows.

The frontend is built in React, rendering a compact 15-minute slot grid that merges contiguous blocks into readable schedule tiles. The UI updates instantly after each solver pass, letting users simulate different weeks, deadlines, or habits.

Challenges we ran into

• Learning multiple new frameworks at once (React UI, OR-Tools, ICS parsing).
• Tuning solver parameters and limits to avoid over-constrained or chaotic schedules.
• Getting timetable UI rendering correct—merging slots, handling spans, and avoiding layout breaks.

Accomplishments that we're proud of

• Building a full scheduling engine from scratch using real constraint programming.
• Successfully parsing and unifying data from Excel, Google Calendar, and Canvas.
• Creating a clean, adaptive React UI that displays minute-level precision without clutter.
• Designing a system flexible enough to handle missing habits, irregular tasks, or last-minute changes.

What we learned

We learned how to integrate constraint programming into real applications, how to normalize messy real-world data, and how to design UI components that remain readable despite high-density information. We also learned how important parameter tuning is—an entire schedule can fail or succeed depending on small adjustments to constraints.

What's next for SleepSync

Next steps include:

• Adding Canvas auto-scraping for assignment detection and effort estimation.
• Integrating geolocation to recommend the nearest study spot during unscheduled gaps.
• Providing personalized analytics for sleep health and study efficiency.
• Letting users generate multiple schedule simulations for exam weeks or work-heavy weeks.
• Expanding into a recommendation engine for long-term planning (degree audits, course balancing).

SleepSync is just the beginning—our vision is a smart, adaptive life-planning system for every student.

Built With

• fastapi
• react