Korean PDF Grammar Checker

Layout-Preserving PDF Proofreading SaaS (V2 Web & V3 Apps in Toss)

Role	Solo Founder / Product Manager & Builder
Context	Independent B2C Micro-SaaS
Stage	Live Service & Monetization (V1 → V2 → V3)
Time	2023.04 – Present

Summary	Expanding a Korean PDF grammar checker SaaS from a V2 web service to a V3 Apps in Toss release.
Scope	Reached 1,500 MAU on V2 web and launched V3 ad-monetized mini-app in Apps in Toss.
Outcome	V2 MAU 1,500 / V3 ad monetization live
Tags	Product, B2C SaaS, Document AI, Monetization
V2 / Web Service	Reached 1,500 MAU on the web service
V3 / Apps in Toss	Monetizing with ads in Apps in Toss

Problem

Capability thesis	Identified a recurring document-workflow bottleneck and evolved a single-purpose utility across three product generations (V1 MVP → V2 1,500 MAU Web Service → V3 Apps in Toss Ad Monetization).
TL;DR	Built a Korean PDF grammar and spell-checking service that inspects exported PDFs without breaking layout/line-wrap artifacts, growing to 1,500 MAU on V2 web and launching V3 with ad monetization inside Apps in Toss.
What	A document proofreading web & mini-app utility that extracts text streams from multi-column/designed PDFs, merges broken line wraps, and highlights Korean spelling/spacing errors in context.
Why now	Job seekers, designers, and marketers finalize resumes, portfolios, and proposals as PDFs in Figma/Canva/InDesign—where native Korean spellcheck is absent and copy-pasting into web spellcheckers corrupts line breaks and character limits.
Problem	When users copy text out of a PDF into a conventional Korean spellchecker, hard line breaks split words in half, producing false errors and forcing tedious paragraph-by-paragraph manual cleanup.

Role	Solo Product Manager, Designer & Full-Stack Builder
Team	1-person product (end-to-end solo execution)
Ownership	Owned PDF text-extraction heuristics, Korean grammar inspection pipeline, V1→V2→V3 UX redesigns, SEO/community distribution, and Apps in Toss platform integration & ad monetization.
Constraints	Zero paid marketing budget, complex Korean spacing rules across PDF coordinate boxes, and strict platform review guidelines for Apps in Toss.

Supported

Job seekers and designers exporting PDFs from Figma/Canva will actively use a tool that checks Korean spelling directly from a PDF file.

Why	Verify why typos repeatedly slip into final portfolio PDFs.
Actions	Observed portfolio review workflows where PDF line breaks caused standard spellcheckers to fail.
Outputs	Identified line-wrap stitching as the core technical enabler.

Why	Ship the smallest working PDF upload + Korean grammar check pipeline.
Actions	Built V1 web MVP with automated sentence merging before grammar inspection.
Outputs	Launched V1 web utility.

Why	Test organic pull without paid acquisition.
Actions	Shared in job-seeker and design communities.
Outputs	Acquired 400+ organic users and confirmed repeat usage during recruiting seasons.

Supported

Improving multi-column PDF parsing accuracy and page-by-page error navigation will drive sustained organic search and repeat traffic.

Why	Resolve V1 parsing edge cases on complex portfolio layouts.
Actions	Rebuilt the extraction engine and UI into V2 with clearer error highlighting and faster response times.
Outputs	Released V2 standalone web service.

Why	Scale monthly active usage organically.
Actions	Optimized search discoverability and error-review UX.
Outputs	Reached 1,500 MAU on the V2 web service.

Supported

Porting the service into Apps in Toss and integrating ad placements will unlock super-app distribution and immediate B2C monetization.

Why	Adapt the web workflow to mobile/mini-app UX standards and integrate ad monetization.
Actions	Built V3 for Apps in Toss, passed platform review, and wired ad placements into the inspection completion flow.
Outputs	Live V3 release on Apps in Toss with active ad monetization.

Problem zoom	Focused on the exact pre-submission moment: users already have a finished PDF and only want to verify 'Are there any embarrassing typos on Page 3?' in under 30 seconds.
Options	(1) Convert PDF to Word first, (2) Plain text-box paste tool, (3) Direct PDF upload with automatic sentence-boundary reconstruction and side-by-side error cards.
Trade-off	Full in-place PDF binary editing is brittle across custom embedded fonts; showing page-mapped error cards with one-click copy/correction kept processing fast, reliable, and privacy-safe.
Final rules	Automatically stitch PDF line-wrap fragments before running grammar analysis, group results by page number, and monetize high-frequency free usage via rewarded/native ads in V3.
Tools	Python, TypeScript, React, PDF.js / PyMuPDF, Apps in Toss SDK
Frameworks	Product-Led Organic Growth, Iterative Version Evolution (V1→V2→V3), Mini-App Ecosystem Distribution

Metrics	V1 validated organic demand (400+ initial users, zero ad spend); V2 web service scaled to 1,500 MAU; V3 launched on Apps in Toss with live ad monetization.
Qualitative outcome	Became a repeat pre-submission utility for Korean job applicants reviewing PDF portfolios and resumes.
Social proof	1,500 monthly active users on V2 web and official approval/release on the Apps in Toss platform.

Learnings	Utility products with episodic B2C usage monetize and retain much better when embedded inside a high-frequency super-app ecosystem (Apps in Toss) rather than relying solely on standalone web bookmarks.
If again	Instrument page-count and document-type telemetry from V1 to segment power users (editors/agencies) vs. seasonal users (job seekers) earlier.

Before	Users manually copied text block-by-block from exported PDFs into 500-character web spellcheckers, fixing broken line wraps by hand.
After	Single-upload PDF proofreading workflow serving 1,500 MAU on web (V2) and generating ad revenue on Apps in Toss (V3).