Park hyobin

Built a documentation automation system for SDK and API docs, using an agent workflow where SDK releases and Jira feedback produce an AI draft checked by editor and user agents, then approved by a writer in Slack. Reduced documentation update lead time from 10 business days to 3, with 50% of AI drafts approved without edits.

Set the roadmap and priorities for four anomaly detection products used by operations and business teams across 40+ games. Wrote PRDs, user stories, functional specs and user-facing policy documents; ran epics and sprints in Jira; covered screen design and QA.

Worked on the Anomaly Detection & Security Solutions PM Team, supporting product development with data analysis and contributing to community issue detection and other data products.

Built CNN models predicting semiconductor thin-film thickness from light reflectance data, achieving best MAE of 0.589. First author of paper presented at IEEE ICCE-Asia 2020.

Gathered client requirements and proposed MVP, phase 2 and phase 3 scope, then trimmed the MVP to the launch date using the development agency's estimates. Resolved a disagreement between client and agency over GPS maps by putting raw GPS collection in the MVP and the map view in phase 2.

Bachelor of Science in Computer Science and Engineering from Korea University of Technology and Education, completed in 2020.