PlanAround

Inspiration

Students often know when an assignment is due, but not how much time they should spend on it or where that work can realistically fit around classes, work, exercise and other commitments.

Calendars show when you are busy, and task lists show what you need to do, but they do not connect the two. PlanAround was built to solve that gap.

What it does

PlanAround turns an assignment brief and a student's actual weekly schedule into a realistic study plan.

Students can add their timetable and other commitments, then upload or paste an assignment brief. AI extracts useful information such as the deadline, module weighting and rubric tasks for the student to review.

PlanAround then estimates how much focused work the assignment deserves and automatically fits study sessions into the student's available time before the deadline.

The generated sessions are also shown directly in the Calendar alongside the student's existing commitments.

How I built it

PlanAround is a web application built with Next.js, React and TypeScript.

AI is used to interpret unstructured information such as timetable screenshots and assignment briefs. The workload calculation and scheduling are handled separately using deterministic TypeScript logic.

This means AI helps understand the input, while the actual planning decisions remain predictable and explainable.

Challenges I ran into

One of the main challenges was getting reliable structured information from assignment screenshots. AI responses can vary, so I added validation and made sure all extracted information remains editable before it is accepted.

Another challenge was scheduling work around real calendar constraints while keeping the generated plan simple and realistic.

I also had to connect the different parts of the application so that an assignment can move from analysis, to workload calculation, to a generated schedule, and finally back into the Calendar.

Accomplishments that I'm proud of

I'm most proud of completing the full workflow rather than building isolated features.

A student can:

enter their real weekly commitments; analyse an assignment brief; review the extracted rubric; calculate an explainable workload; generate study sessions around their availability; see those sessions in their Calendar.

The system also keeps AI separate from the workload and scheduling logic, making the final plan easier to understand and trust.

What I learned

I learned that AI is most useful here for interpreting messy information rather than making every decision.

Using AI for extraction and deterministic logic for workload and scheduling produced a system that is both flexible and explainable.

I also learned how important it is to design around a complete user workflow rather than adding as many features as possible.

What's next for PlanAround

The next step would be to make scheduling more personalised.

Students could choose preferred study times, avoid early mornings, set daily study limits and regenerate plans when their availability changes.

Future versions could also include calendar integrations, automatic replanning, notifications and support for managing several assignments together.

Built With

• cloudflare-workers
• featherless-ai
• github
• localstorage
• next.js
• react
• tailwind-css
• typescript
• vercel