At TikTok Shop, I analyze user feedback, rejection labels, and case-level data to identify recurring seller support issues and turn them into product and process improvements. That work increased rejection-related CSAT by 9.8pp.
I evaluated AI/LLM decision quality using labeled datasets, case audits, and false-positive analysis, collaborating with Algorithms and Operations to improve model accuracy to 98%. I also used AI-assisted clustering and root cause analysis to reduce issue diagnosis time by 50%.
At TikTok Shop, I planned go-to-market and product rollouts for seller self-service features, aligning launch readiness with support workflows and post-launch monitoring. Feature adoption increased by 10pp.
At Jamesdar, I led development and launch of a direct-to-consumer e-commerce platform, translating customer needs into product requirements and feature improvements. Conversion rate increased by 0.8pp, while product and content improvements reduced bounce rate by 15.2pp.
