Inspiration
In a tailoring workshop, choosing the correct size is not always a simple one-size answer. A child’s chest and shoulder measurements may correctly match more than one garment size, while parents often struggle with technical sizing tables.
AlbaFit turns practical tailoring knowledge into a clear and accessible sizing assistant.
What it does
AlbaFit uses two key body measurements:
- chest circumference
- shoulder-to-shoulder measurement across the back
It compares them with real sizing ranges used for children’s long albs and shows every technically valid size.
When two sizes fit, AlbaFit clearly explains which option gives a closer fit and which provides more room. Height and additional measurements can be used as supporting checks.
All measurements remain in the browser and are not stored.
How we built it
The project combines real workshop sizing data and many years of tailoring experience with OpenAI Codex and GPT-5.6.
Codex helped:
- analyze the original sizing spreadsheet
- translate Polish tailoring terminology into internationally understandable language
- implement the sizing logic
- create a responsive browser interface
- build regression and data-integrity tests
- simplify the experience for parents and non-technical users
The application is built with HTML, CSS and JavaScript and is hosted publicly with GitHub Pages.
Challenges
The main challenge was preserving professional tailoring logic without overwhelming users with technical terminology.
Another challenge was representing situations where several sizes are equally valid. Instead of forcing a false single answer, AlbaFit separates technical fit from the user’s preference for a closer or roomier garment.
Accomplishments
AlbaFit transforms workshop knowledge into a working, privacy-first public prototype.
It provides transparent recommendations, explains why a size fits and keeps professional details available without cluttering the main interface.
What we learned
AI works best when technical capability is combined with real domain expertise. Codex made it possible to rapidly turn practical knowledge, sizing data and production experience into a tested application.
What's next
Future versions could support additional garments, historical order analysis, measurement anomaly detection, workshop inventory and batch production planning.
Built With
- codex
- css3
- github
- gpt-5.6
- html5
- javascript
- openai
Log in or sign up for Devpost to join the conversation.