ZHEN(Steven) ZHAN

Manage 3 teams of 28 engineers/managers in Amazon Ads Demand Side Platform (DSP):

• Own core infra and products of Amazon's fast growing $40B+ advertising business (ads indexing, federation platform, contextual targeting, and creative optimization)
• Leading Amazon DSP top initiative (~$2.5B revenue opportunity) to transform monolithic ad serving system into federated services and enable modeled based targeting with budget optimization. It is projected to increase customer feature development speed by 9X; while maintaining support of more than 10 million QPS, less than 100ms latency, and 99.9% availability
• Building next generation of Cookie-less targeting platform to launch a variety of ML powered contextual products. Launched products have generated north of $120MM revenues in less than one year
• Growing the team from 3 engineers to a distributed org and collaborating cross-functionally with product, science, analytics, finance, and legal
• Day to day functions include building, managing, and coaching teams, managing stakeholders, defining team vision and executing on product roadmap, and ensuring engineering excellence

Managed cross-functional team that was responsible for building trust platform to detect and prevent user abuses and fraudulent activities for Grab (Uber of Southeast Asia), which provided everyday services to more than 100 million active users:

• Worked with multiple business verticals (Food delivery, Ride hailing, Payments, Parcel delivery, etc) and multiple stake holders (business owners, PMs, DSs, DAs, policy owners and engineers) across multiple countries to build platformized solution that drove down fraud losses for entire Grab by 10 folds while user bases jumped more than 20 times and business complexity increased significantly
• Built and grew engineering team from scratch while also leading multiple cross functional and cross country (US, Singapore, China, and Malaysia) feature teams to achieve various business initiatives
• Architected real time fraud detection as a data intensive system with more than 50 gRPC APIs and 20 clients that supported more than 50k QPS on critical paths while handling challenge of double QPS every half year and constantly changing fraud patterns

Led team to develop next generation of organic user acquisition platforms utilizing machine learning and big data technologies:

LeEco Global Search Marketing/Growth Platform (SEO) - to grow organic search traffic for all business units of the company

• Architected business requirements, technical design, project planning, and testing for all services on platform; collaborated with 11 teams from 3 countries (US, China, and India); actively implemented four key services
• Key results from #1 internal client - le.com (Chinese Youtube/Netflix - 15 mil DAUs): 3k daily automated and optimized video landing pages - CTR 22%, average targeted keywords Baidu/Google ranks 1.8, overall search organic traffic improves 20%; internal search engine CTR improved 18%; internal traffic churn rate and engagement rate improved 5% and 3% respectively

Hot Topic Detection System - to detect trending online topics in real time.

• Contributed in same job roles as for the SEO projects
• Key services/components included template based web crawler, machine learning based NLP service, machine learning based trend detection service, real time data pipeline, and A/B testing integration
• Topics were captured within one minute, trends were detected within an hour; false positive rate 42%, false negative rate 15%; significantly improved performance of downstream clients like SEO, SEM, Recommendation, and human editors

Complete engineering role with focus on development and additional job responsibilities included project design, testing, deployment, business monitoring, and on-call support