Lugoffice vs DocLens

The hard part of a meeting isn't the meeting — it's the paperwork after. Lugoffice turns a spoken Cantonese meeting into polished written documents you can hand in right away. Lugji captures what was said; Lugoffice gets it done. Seven document types from one conversation · Meeting Minutes — summary, discussion, decisions, action items · Announcement — title, audience, body, sign-off · Formal Email — subject, greeting, body, courteous tone · Report — a structured report with summary and sections · Decision Notes — only the decisions actually made, with rationale and owners · Polish — proofread, tighten or expand your own draft (wording only, nothing invented) · Weekly Report — combine a week of meetings and notes into one review Cantonese-first Built for spoken Cantonese → written Chinese; also handles Mandarin and mixed Chinese/English. You choose what's sent, and only the text Your recording never leaves your device (it stays with Lugji) — Lugoffice only reads text, never your audio. Before every generation you see exactly what will be sent, and confirm first. Text is processed by enterprise Google Vertex AI and is not used to train AI models. Documents are saved to your own iCloud or Obsidian. Lugbox for organizations (licence required) Organizations running a Lugbox can send the documents you mark over their internal network to their own Lugbox, for local AI search and an organization shared library. Enquiries: sales@lugji.app Works best with Lugji (Cantonese meeting recording + transcripts) — it reads your Lugji meetings directly. No Lugji needed either: paste any text and still get documents. iPhone and iPad · iOS 18.

DocLens is a fully offline frame-select OCR tool: draw a box, extract exactly the text you circled, and translate it on your device. The liability clause in a contract, the date on a court ruling, one verbatim line from an interview: you choose the block, it reads only that. Headers, footers, ads and noise are never touched. Key features: 1. Frame-select OCR: shoot or pick a photo, drag a box, and only the text inside is recognized. Low-confidence lines are dashed-underlined for line-by-line proofreading before translation. 2. Accurate recognition: selectable recognition languages with automatic per-line detection for mixed Chinese and English; handwriting and rotated text supported; Enhance and Retry rescues poor scans. 3. On-device translation: results are translated paragraph by paragraph for fluent output. Chinese, English, Japanese, Korean, French, German and Spanish are supported, and language packs can be downloaded in advance. 4. QR and barcode scanning: frame a code to open URLs in an in-app browser, dial numbers, compose emails, open locations, or copy Wi-Fi passwords and contact fields. 5. Fully offline: recognition, translation and barcode scanning all run on your device. No account, no analytics, no cloud upload, and no network access from shutter to text. 6. Minimal permissions: DocLens never asks for full photo library access, only permission to save your exports. Data stays in the local sandbox and database. 7. Torch and Enhance Retry: turn on the light for close-up documents, and re-run recognition with boosted contrast when a scan comes out poorly. Membership: New users get a free 7-day trial, extendable once by another 7 days from the membership page. After the trial, you can still scan free once per day. VIP is a one-time purchase: no subscription, no ads, no tracking. Great for contract review, hearing and meeting notes, interview transcription, research excerpts, receipt and form data extraction, and translating foreign signs. Your files go nowhere. That is what makes DocLens different: it does not make decisions for you. It hands judgment back to you, and keeps privacy structural with on-device computing.