Archery Note

Inspiration

I am a member of my university’s archery club. Because archery is a niche sport, I felt that athletes have very few trustworthy software tools to support their daily practice. I started Archery Note to build the kind of tool I wanted to use as an archer.

What I built

My main focus was bringing together information from several angles—arrow grouping, scores, and body movement during each shot—so that an archer can understand the day’s practice and reflect on it at a glance.

Archery Note provides phone-first score recording, grouping analysis, form records, equipment and sight notes, and a personal growth timeline. It compares an athlete only with their own previous practices. Suggestions show the recorded reason behind them and lower their confidence when there is not enough data.

All records and analysis stay on the device. The app works as an offline PWA and supports local snapshots, JSON backup and restore, and CSV export.

UI design

The most difficult part was designing the interface while simulating how an archer would actually use it during practice: preparing equipment, recording arrows quickly, checking grouping, and reviewing the session afterward.

I wanted the recording flow to remain simple even though the app combines several kinds of information. Detailed analysis is therefore placed after the recording flow instead of adding more controls to the main screen.

What changed during Build Week

• Added an explainable growth dashboard
• Added fixed 7, 30, and 90 day views
• Added next-practice suggestions with visible evidence and confidence
• Added clearly fictional demo records that can be removed separately
• Improved first-load performance and release verification

How I used GPT-5.6 and Codex

I used GPT-5.6 in Codex to inspect the existing architecture and Git history, design pure analysis functions, implement the dashboard, add regression and Playwright tests, review privacy boundaries, and verify the release.

I made the product decisions: which information should be combined, how an archer moves through the app, how simple the recording screen should remain, and why practice data must stay on the device.

What I learned

I learned that useful practice analysis needs more than a score. Grouping, form, equipment, and changes over time need to be read together. I also learned that a recommendation should explain its evidence instead of presenting an unexplained answer.

What’s next

I want to test Archery Note with more real practice workflows and improve its longitudinal form analysis while continuing to keep athlete data private and local.

Built With

• javascript