Skip to content

WIP_REPORT.md

Latest commit

 

History

History
74 lines (52 loc) · 3.35 KB

WIP_REPORT.md

File metadata and controls

74 lines (52 loc) · 3.35 KB
Raw

Restock – Work-in-Progress Report

1. Problem Statement

Small convenience stores still use paper, clipboards, or basic spreadsheets to track daily inventory. These manual workflows are time-consuming (40–60 seconds per item), error-prone, and often skipped entirely due to workload. The goal of Restock is to replace the repetitive “look-→ write-→ update Excel” workflow with a lightweight, offline-friendly OCR-powered scanner, saving time per entry while improving record accuracy, accountability, and ease of hand-off between shifts.

2. MVP Scope (✅ = done)

Feature Status
OCR via Google Vision – scan to text ✅
Camera UI with live preview ✅
Text → Log Entry parser ✅
Editable list (inline quantity) ✅
Manual-entry fallback ✅
Delete / duplicate actions in list ✅
Offline local-storage (AsyncStorage) ✅
TestFlight build ✅
Basic UX polish (icons, spacing, feedback) 🔜

3. Metrics Being Tracked

Metric Baseline (manual) Current WIP Target (v1)
Scan-to-log time ~40–60 s per item ~4–5 s per item* < 3 s per item
OCR Accuracy (labels) N/A (manual typing) ~80%* > 90%
Offline Reliability N/A Works offline Add “pending sync” UX

*Based on internal manual tests with dummy labels.

4. Upcoming 7-Day Focus (Iteration sprint)

Priority Task Description
P0 Optimize state updates + debouncing to reduce scan latency
P0 Improve text cleaning and basic fuzzy-matching to increase tag accuracy
P1 Fix list UI overflow (scroll patch when > 5 items)
P2 Start collecting feedback from small store clerks via TestFlight (1-2 testers)

5. Beyond All Versions: Future Roadmap (Only if I want have store permission)

  • Product DB integration for common SKUs / barcode lookups (If my store gives permission)
  • Auto-classification of items by category (dairy, snacks, etc.)
  • Export / sync options → Google Sheets, Shopify POS, WhatsApp reminders
  • Session history tab: group scans by date/time
  • Staff audit view → “who added what, when”

6. Notes

  • Demo screenshots and demo video use random dummy text tags
    (e.g., “Apple”, “Banana”, “ELEVEN”) to avoid legal/trademark complications.

  • Designed and built using: React Native (Expo), TypeScript, Google Vision OCR, EAS builds.

Tech Stack

  • Framework: React Native (Expo + TypeScript)
  • OCR: Google Vision TEXT_DETECTION
  • Storage: AsyncStorage (offline cache)
  • CI/CD: EAS builds → TestFlight

Last Updated: <Fill date before submitting>