sinye.site

Projects

How I understand — and reduce —
the complexity people face.

Two public projects: Dear Me and TonePilot. Each case goes deep on the problem, alternatives, judgment, user control, and evidence.

Product planning sketched on a whiteboard
Paper documents and a pen spread on a desk
Dear Me Prototype

Statements scattered across texts, paper, and email — organized into one ledger, sources intact.

AI reads and structures statements from every channel, keeping the original, its confidence, and edit history, so users verify and confirm.

  • User problemRe-copying statements every month, worrying about missed bills.
  • RolePlanning · Information architecture · Product design
  • StatusPrototype — validating working screens · Updated 2026-07
  • EvidenceField schema · User flows · Review screen design · Data boundaries
Try the app
7월 지출 2,149,180원

이번 달 요약

쇼핑1,494,380원
생활600,000원
식비30,900원
✎ 직접 적기📷 영수증 찍기

최근 내역

쿠팡(주) 쇼핑 · 7.27-7,400원
테스트상점 기타 · 7.26-2,000원
월세 생활 · 7.26-600,000원
Playing an electric guitar on stage
TonePilot Prototype

Describe the tone you want — get a gear-aware multi-effects setup with reasons.

Based on your gear, playing environment, and desired sound, it builds a usable starting chain and explains "why this value" in a player's language.

  • User problemPresets sound different on my gear, and I don't know where to start adjusting.
  • RolePlanning · AI workflow design · UX
  • StatusPrototype — validating working screens · Updated 2026-07
  • EvidenceDevice schema design · User flows · Explanation UI · Safety principles
Try the app
후크 (예술이야 반복)1:40 → 1:57

짧고 강렬한 후크 파트, 하이게인 유지하며 딜레이/리버브 강조.

AMP · Gain7/10하이게인 유지
DLY · Dual DelayTime 450ms리듬 악센트 강조
REV · SwellDecay 40%악센트마다 스웰감

👉 P04 프리셋 호출, 픽업 1번 포지션 유지.

Comparison

Two projects, one philosophy

Different areas of life, one message: translating complexity into understandable flows.

AspectDear MeTonePilot
Source of complexityDifferent channels, formats, items, billing datesSignal chains, gear combinations, endless parameters
AI's roleExtraction, structuring, categorizing, duplicate candidatesChain & parameter starting points, feedback-driven revisions
User's roleVerify sources, edit, merge, delete, confirmJudge the sound, give feedback, adjust, save
Core trustNever lose the original source or change historyExplain why this setting, and flag gear constraints
Research linkPersonal information, document understanding, privacyHuman-AI co-creation, preference learning, explainable recommendation

Product Principles

Principles both products share

Sources

AI results must always trace back to their original source and evidence.

Explanation

Explain why a suggestion was made, in the user's language — not just the result.

Right to correct

Users can always edit, merge, delete, and undo.

Privacy

Users hold the rights to store, delete, and export their own data.

Compatibility

State the supported range and constraints; never invent missing features.

Honest status

Concept / Prototype / In development / Shipped are labeled consistently.

Note

Planned research projects (baseball player growth prediction, K-pop pattern analysis) are published with their status on the Research page, not here.

Curious how the two cases prove their approach?
Read the full case studies.

Try Dear Me Try TonePilot