• Spearheaded the 0-to-1 development and global GA of TikTok’s full and semi-automated ads targeting products, successfully driving 20% cost penetration across the advertising ecosystem.

• Architected and modernized the Lookalike (LAL) platform, significantly enhancing system scalability and reliability while expanding its application scope to drive enterprise-wide adoption.

• Led a strategic compliance-driven transformation of the LAL platform, navigating complex global legal requirements to mitigate risk while simultaneously optimizing algorithms for sustained revenue growth.

• Championed engineering excellence by re-engineering core infrastructure and deployment pipelines, drastically reducing operational overhead and fostering a culture of high-reliability product development.

Lead engineer of Amazon Advertising Bulk Tool

• Modularized and refactored the Bulk Tool product to increase its code coverage from <30% to >90%, which made the product much easier to maintain and greatly increased its quality.

• Upgraded the Bulk Tool to support multiple ad programs and improved its ease of use, which drove its usage by 5 times.

• Designed and implemented Amazon Advertising Service JAVA SDK for both internal and external users ahead of schedule, the SDK provides significant convenience for Amazon Advertising API consumption.

• Migrated Bulk Tool's backend to Amazon Advertising Public API Service to streamline Amazon Advertising's technology stack.

Lead engineer of Amazon ads rendering data pipeline enhancement

• Built a data pipeline to publish KPIs data into ElasticSearch cluster to enable business owners optimize Amazon advertising marketplace in the real time.

• Integrated alarming and visual diagnostic functions into the data pipeline to enable Op Team monitor ads rendering pipeline end-to-end with ease.

Standardized WW Amazon ads placements to increase productivity

• Designed and implemented the next generation of ads delivering pipeline across Amazon desktop, web, mobile app to increase ads coverage and relevance.

• Unified the front-end solution, including CSS styling and Ad Feedback design, for ads on mobile and desktop, which drastically reduced the maintenance burden.

• Optimized client-side javascript code with error handling logic, which reduced ad javascript fatals from hundreds per hour to tens per day.

• Implemented a SafeFrame solution to support third-party ads on Amazon website without risk of broken customer experience, which directly generated over $20M ad revenue.

• Upgraded Amazon Ad Placements client-side solution to measure the ad viewable rate to support a viewable CPM (vCPM) bidding model.

Migrated all Amazon O&O desktop ad placements to next generation of header-bidder architecture

• On-boarded multiple new ad programs onto Amazon O&O ad placements, which drove ad revenues up via bidding competitions.

• Updated front-end code to support client-side call to ad servers to retrieve ads, which provides the flexibility to satisfy different latency requirements.