• Designed and implemented a Text-to-SQL solution leveraging LLMs and a curated semantic layer, significantly enhancing productivity in data analysis.• Developed an enhanced interface for curated data access, optimizing discovery and access to data warehouse systems.• Constructed a data processing framework for near real-time alerts, integrating technologies like StarRocks, Flink, Spark, Hive, and Airflow.• Established and managed big-data pipelines using Spark, aimed at marketing performance evaluation and fraud detection.• Directed the creation and implementation of Sage, a platform dedicated to analytics and evaluating growth.• Created real-time monitoring dashboards and systems for detecting anomalies in email and push notification campaigns.

• Assumed a leading role in numerous projects, including setting OKRs/KPIs, overseeing task distribution, and handling requests for new features.• Engineered and deployed data processing and machine learning services tailored for financial forecasting applications.• Developed a range of microservices using Python, Java, Golang, and UberFx to meet essential business and product requirements.• Spearheaded the creation and launch of a real-time driver incentive processing system, resulting in a 10,000-fold reduction in payout latency compared to the previous weekly batch system.• Worked in tandem with data scientists to fully integrate machine learning technologies for automated campaign generation.• Initiated and enhanced the bulk load method for uploading near-time incentives to Cassandra, achieving a 60-fold increase in efficiency compared to previous methods.

• Oversaw the full-cycle development of a new data warehouse system for LinkedIn's Help Center.• Transitioned the Voice of Member and Customer system from weekly batch processing to a real-time processing framework using ElasticSearch.• Executed the creation of several ETL data pipelines, utilizing tools like Hadoop, PIG, Spark, Java, Python, Azkaban, AppWorx, MySQL, and Teradata.• Collaborated closely with the Business Analytics team to develop various functionalities for the Data Mart for Data Mining (DM2).• Explored and refined different methods to enhance data pipeline performance for machine learning feature engineering on Hadoop.

• Led the establishment of a remote R&D team for Bing Satori Knowledge Graph in Suzhou, China, in the capacity of a Tech Lead.• Took charge of designing and creating the self-serve infrastructure for Bing's knowledge data reporting platform.• Developed numerous essential features for monitoring Bing’s service availability and latency.• Constructed comprehensive data processing ETL pipelines and workflows using the Data Warehouse Controller platform, meeting stringent reporting SLAs.• Worked in conjunction with data analysts to enhance the performance of the Bing search engine and resolved numerous live site issues.• Employed various methods to improve the efficiency of Cosmos virtual clusters and data processing pipelines.

• Played a pivotal role as a founding member of adCenter (later Bing Ads), where I developed the initial BI and Billing ETL pipelines for advertising functionalities.• Progressed to lead the ETL data pipelines team, expanding it to approximately ten members for the Yahoo/Bing integration.• Created various BI applications to produce reports for both advertisers and publishers, handling large volumes of raw web data.• Engineered automated testing and validation applications using C#/.NET for data processing systems.• Led the transition of legacy ETL applications to MapReduce-based distributed systems, managing the daily processing of terabytes of data.• Coordinated with database teams to establish efficient and reliable data distribution, extraction, and loading interfaces.• Developed a thorough BI reporting and OLAP system for Mobile Advertising using the SQL Server Business Intelligence Suite (SSIS/SSAS/SSRS).

• Crafted graphic user interfaces (GUI) for Windows client applications utilizing Visual C++/MFC.• Upgraded and managed a third-party MFC software package, Ultimate Grid, widely employed in various data entry GUI applications.• Contributed to the development of the IWR (Interactive Web Responder) website, designed for health insurance participants to access information online, using technologies like J2EE, Apache, and BEA WebLogic/Portal.• Engineered an FTP file transfer monitoring system and a Java FTP Client for automated electronic data processing.• Created DLLs enabling Java applications on Windows to interact with applications in the Tuxedo domain on UNIX/VMS systems.• Developed middle-tier service routines facilitating the connection between MFC client applications and the Oracle database server.