Memoli is an iOS application designed to act as a mindful assistant for modern living and sustainability. By tracking the shelf-life of household items, calculating Period After Opening (PAO) alerts, and checking ingredients against health and safety standards, the app helps families eliminate "sunk cost" waste and protect their loved ones. Built as a collaborative group project starting in February 2026, Memoli blends on-device vision text recognition with a full-stack data synchronization layer to ensure a seamless, real-time shared cabinet view for all family members.

Problem

In busy urban households, managing cabinets and storage spaces often leads to clutter. Products are frequently forgotten until after their expiration dates, resulting in both environmental waste and financial loss (sunk cost). Additionally, family members often purchase duplicate items because they lack visibility into what is already in-stock at home. Furthermore, manually reading and verifying complex ingredient lists on product labels for safety and compliance with health standards is a tedious, time-consuming process.

Solution

Memoli transforms household management into a shared, collaborative, and intelligent experience. By leveraging on-device camera scanning and optical character recognition (OCR), users can digitize products and catalog shelf-life instantly.

For safety, the app analyzes scanned ingredients against safety indices to highlight potentially harmful components. To coordinate the household, a real-time fullstack synchronization backend keeps family members updated on cabinet inventory. Automated push notifications alert users when items are nearing their expiration or PAO limit, preventing waste and ensuring that everyday essentials are fresh, safe, and cost-effective.

Features

Smart Scan, Zero Effort

Users can skip manual text entry. By leveraging on-device vision, Memoli scans product labels to instantly capture titles and descriptions, making inventory cataloging fast and error-free.

Proactive Expiry & PAO Tracker

The app calculates use-by dates and Period After Opening (PAO) intervals. It triggers automated notifications before products expire, helping families stay healthy and consume mindfully.

Ingredient Safety Guard

Memoli extracts ingredient lists from product packaging and matches them against health and safety standards. This provides instant alerts on potential allergens or toxic substances, guiding users to healthier options.

Family Synchronization

Synchronize updates across multiple devices in real time. Memoli maintains a shared family shelf, preventing duplicate shopping and ensuring everyone knows what is available in the cabinet.

Architecture

Memoli is designed with a client-server fullstack model. The system comprises a native iOS app built in Swift/SwiftUI for local intelligence and on-device machine learning, integrated with a lightweight Node.js/Express.js backend for user coordination, and a Next.js web application for product discovery.

Native Client (Swift & SwiftUI): Manages high-performance camera capture, local OCR scanning via Apple's Vision framework, and provides a sleek, responsive user interface.
Backend REST API (Express.js & Node.js): Serves as the synchronization service, handling family groupings, inventory updates, and real-time database state propagation.
Web Interface (Next.js): Provides a clean web representation, serving as a landing portal and allowing users to view product indices and features.

Tech Stack

Swift & SwiftUI: Used to build the native iOS app, exploiting Swift's type-safety and SwiftUI's declarative structures for fluid interfaces and high-performance camera processing.
Express.js & Node.js: Powers the API layer for authentication, family synchronization, and database storage.
Next.js: Used to build the landing pages and admin panel.

My Contributions

iOS App Design Revamp

I spearheaded the overhaul of the native iOS app by adopting and implementing a completely new design system, resulting in a cleaner user experience, smoother micro-interactions, and better visual consistency.

Ingredients Database Schema Redesign

Redesigned the database schema for the ingredients safety checker. This optimized structure supports complex ingredient relationships, standardizes naming variations, and improves index query performance during scans.

Crowdsourced Ingredients OCR Pipeline

Developed a scheduled task pipeline that allows users to upload raw ingredient label images. This task automatically processes uploads by extracting text using OCR, normalizing chemical names, and aggregating safety frequencies. The pipeline outputs data candidates for admin approval, ensuring a safe and verified database growth mechanism.

Landing Page & Web Support

Contributed to the development of the Next.js landing page, focusing on minor layout tweaks and mobile responsiveness to support the initial public announcement of the app.