AI Mobile Product

IngRyn

An AI-powered mobile product focused on rapid ingredient scanning and intelligent analysis workflows.

01 / Type

AI Mobile Product

02 / Role

Product Engineer

03 / Stack

React Native • Expo • Supabase • Gemini AI • OCR Pipeline • Mobile Camera Workflow

04 / Status

Active Development

View Source
IngRyn screenshot 1
Screen 01
IngRyn screenshot 2
Screen 02
IngRyn screenshot 3
Screen 03
IngRyn screenshot 4
Screen 04
IngRyn screenshot 5
Screen 05
IngRyn screenshot 6
Screen 06
01

Overview

An AI-powered mobile product focused on rapid ingredient scanning and intelligent analysis workflows.

02

The Problem

Understanding product ingredients on the go often requires manually reading dense labels and researching unfamiliar ingredients one by one.

03

The Solution

Created a mobile scanning workflow that captures a product label, extracts ingredient text through OCR, and sends structured information through an AI analysis layer before presenting the result to the user.

04

Architecture

[ Mobile Camera ] │ ▼ [ OCR Processing ] │ ▼ [ Ingredient Extraction ] │ ▼ [ AI Analysis Layer ] │ ▼ [ User Insights ]
05

Engineering Decisions

React Native + Expo

Used Expo and React Native to iterate quickly on a cross-platform camera-driven product while retaining access to native capabilities.

Supabase Backend

Used Supabase for persistence and authentication, keeping application data and user state behind a managed backend.

OCR → AI Pipeline

Separated image capture, text extraction, ingredient processing, and AI analysis into distinct stages so each part can be improved independently.

06

Challenges

Challenge 01

OCR Accuracy

Ingredient labels vary in lighting, typography, curvature, and image quality, making preprocessing and extraction quality important to downstream analysis.

Challenge 02

Mobile Responsiveness

Balanced camera capture, image processing, and network requests so the interface remains responsive during longer-running analysis.

Challenge 03

AI Request Reliability

Handled loading, failure, and retry states around AI requests to account for variable mobile network conditions.

07

Future Improvements

  • Verified static ingredient database
  • User-driven personalization and ingredient preferences
  • Robust offline support
  • More deterministic analysis and validation layers
08

Impact & Outcomes

  • 01AI-powered mobile workflow
  • 02Camera-based interaction
  • 03OCR + AI integration
  • 04Structured analysis experience