Chung-Chia Kuo

B2C – AI Music & Music Video Generation App

• Owned the music video flow from day one, from storyboard editing to generation and results, shipping it in 10 languages within 1.5 months using Next.js, TypeScript and Tailwind CSS.
• Turned AI-built prototypes into production code with Claude Code, and used it across the cycle to implement specs, write tests and fix bug tickets.
• Validated photo and audio uploads in the browser, enforcing the single-face rule with MediaPipe and backend length limits with an audio trimming flow.
• Corrected flow and UI errors in the product spec, and fixed the Countly and GA4 tracking specs against the real user flow while implementing them.
• Fixed a TanStack Query bug that left finished jobs polling the server forever, and added regression tests to prevent it from returning.

B2B – AI/AR Beauty Tech Solutions

• Delivered React virtual try-on and skin analysis UIs to 30+ global beauty brands, integrating an in-house AI/AR SDK on photos and live video.
• Extended the SDK's client-facing API with init parameters and events for consent, recommendations and detection results, and documented them for client developers.
• Customized and localized brand experiences for up to 15 markets per brand, including survey, skin report and product routine pages for a skincare analysis flow.
• Implemented consent and privacy flows for client markets, including BIPA biometric consent and cookie-based analytics opt-out.
• Resolved 300+ client bug tickets in 14 months, tripling monthly throughput (18 → 54) after building a Claude Code + n8n pipeline that triages bugs and auto-fixes simple UI issues for my review.

• Led 4–5 person TA teams for Programming Design (C++) and Operations Research, designing test cases and grading programming assignments.
• Ran weekly review sessions and office hours for Programming for Business Computing (Python), serving 800+ students per term across three semesters.

• Built and deployed a real-time anomaly alerting system for an industry partner's light sensors, serving a pretrained ML model through a Flask and MongoDB backend on Google Cloud Platform, with alerts via the LINE Messaging API.