Readplace

The Problem

Pocket shut down. Omnivore shut down. Read-it-later users keep getting burned: services die, exports go missing, and reading data ends up locked behind opaque AI training pipelines. Anyone who wants a quiet place to save articles and actually read them later is left choosing between abandoned tools and bloated newsreaders that monetize their attention.

The Build

Readplace — a lightweight read-it-later app built around a 10-year personal reading system, with privacy and data portability as first principles.

The core flow is deliberately small: paste any URL, get a clean reader view plus an AI-generated TL;DR in seconds. No signup required to read. Saving uses a one-click browser extension (Firefox + Chrome). Articles are extracted with Mozilla's Readability engine, PDFs are run through vision OCR, and summaries come from DeepSeek V3.2. OAuth tokens are stored locally; data export is a single JSON file you own.

Tech stack: Mozilla Readability, DeepSeek V3.2 (summaries), Deep Infra (PDF vision OCR), browser extensions for Firefox and Chrome, source on GitHub.

Key features:

• One-click save via Firefox / Chrome extensions
• Clean reader view + TL;DR summary on every saved article
• Public reader view — no account needed to read a saved link
• PDF support with OCR for image-based pages
• Mobile web reader
• Local OAuth token storage, JSON export — your data, your file
• Built in Australia, solo-developer maintained

Results

• Live at readplace.com, serving as a daily reading tool for early users
• $3.99/month subscription to cover infrastructure; first 50 users get lifetime free access
• Built and maintained solo by Fayner Brack, creator of js-cookie (~22B downloads/year) — a one-person operation with no VC pressure to monetize attention
• Demonstrates how a focused, single-purpose tool can fill the gap left by shuttered incumbents without trying to become a platform