Ivory
AI OCR platform for infectious-disease childcare services
2nd Place - kakao × goorm Hackathon 2024
Overview
Working parents of children with infectious diseases must navigate heavy paperwork to access Korea's government childcare support. Ivory automates the entire journey: photograph a doctor's note and school-absence certificate, get an instant eligibility verdict from AI OCR, see your government subsidy auto-calculated by income bracket, pay only the copay via KakaoPay, and get matched with a verified caregiver - with real-time status tracking. I served as PM of the 6-person team and contributed to the React frontend (splash, start, and login screens).
Demo
Gallery
Highlights
AI OCR document validation
Naver CLOVA OCR with two custom document templates (medical certificate, absence confirmation) extracts diagnosis, infectiousness, and onset date - returning an instant eligibility verdict instead of a days-long manual review. Chose CLOVA over Google Vision for Korean-document accuracy and template support.
Subsidy calculation & KakaoPay
Income-bracket rules compute the government subsidy automatically; parents pay only the remaining copay through KakaoPay in one step.
Caregiver matching & live tracking
Verified caregivers are auto-matched on application, with a four-stage status flow (applied → matched → caring → complete) pushed in real time over STOMP/WebSocket.
PM of a six-person team
Led planning and delivery across design, FE, and BE under hackathon pressure - and shipped frontend screens myself. The team took 2nd place among the kakao × goorm cohort.