WorthHours

• 

  Home screen

• 

  Summary

• 

Inspiration## Inspiration

WorthHours started from a simple problem: shift workers often know when they worked, but it is harder to keep track of hours, breaks, overtime, expected pay, appointments, and notes in one clean place.

The goal was to build a small, practical shift calendar that feels fast on a phone and keeps the important numbers visible.

What it does

WorthHours lets the user track work shifts, breaks, overtime, expected pay, appointments, and notes in a calendar.

There are two ways to enter data:

• From the calendar, by choosing a date and adding a shift, template, or note.
• With AI Quick Add, by writing a sentence in natural language.

AI Quick Add uses GPT-5.6 to convert text into structured calendar drafts. The user reviews the preview before anything is saved.

After confirmation, WorthHours uses its own app logic for calculations, warnings, and saved calendar data.

How we built it

The app is built with Codex , and deployed on Vercel.

For the Build Week AI layer, we added a small serverless endpoint that calls the OpenAI API. GPT-5.6 returns structured data, not final calculations.

Codex helped build and test the endpoint, the structured output schema, the preview flow, warning logic, README, and submission materials.

What we learned

The most important design decision was to keep AI away from the payroll math.

AI is useful for understanding flexible human input. But the app should remain responsible for the business logic: worked time, breaks, overtime, pay rules, and warnings.

That keeps the experience useful without making the result feel unpredictable.

What's next

After Build Week, the AI layer can become the foundation for a future premium version.

The public version can stay focused and simple, while the AI workflow can continue to improve with better parsing, more entry types, and richer review tools.

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for WorthHours

Built With

• codex
• css
• github
• gpt-5.6
• javascript
• openai-api
• react
• serverless
• typescript
• vercel
• vite