CASE FILE № 010 — CaptureNote

CLIENT OR PRODUCT

Independent product by Sidi Li (capturenote.io).

MY ROLE

Independent product builder — web + iOS, end to end

THE PROBLEM

Students and knowledge workers capture screenshots constantly, but text and structure stay locked in images and hard to reorganize later.

CONSTRAINTS

• •Cloud multimodal AI on web vs local-first privacy on iPhone
• •Visual structure — tables, diagrams, formulas — not flat OCR only
• •Keep source image, structured note, terms, and Markdown together
• •iOS Shortcuts, folders, and search for a durable library

THE INVESTIGATION

Mapped the real loop: capture → understand structure → summarize → organize → find again, across web speed and on-device privacy.

THE STRATEGY

Ship a dual surface: fast multimodal AI on the web for insight, plus a native SwiftUI iPhone app for local-first library, Shortcuts, and offline notes.

WHAT I BUILT

• •Web capture flow with multimodal AI note generation
• •Title, summary, key points, terms, and clean Markdown output
• •iPhone local library with folders, search, and source retention
• •Shortcuts actions for screenshot-to-note workflows

THE HARD PART

Keeping visual understanding trustworthy while offering both cloud speed and local-only privacy without fracturing the product story.

THE VERDICT

• •Live product site at capturenote.io
• •Independent web + iOS product owned end to end
• •Capture-to-note path designed for study and knowledge work

TECHNICAL EVIDENCE

SwiftUI · iOS · Web · On-device OCR · AI · Shortcuts · Education