2024 — Now
San Mateo, California, United States
SQL team
San Jose, California, United States
Engaged in a pivotal role within the core performance optimization project, successfully achieving a reduction in memory usage for the simulator procedure through proficient use of C and C++.
Played a key part in ensuring the sustained business collaboration with major clients, showcasing a commitment to enhancing efficiency and fostering strong customer relationships.
Automated routine compiling and building tasks with Shell scripting, leading to reduction in manual workload.
Helped the performance optimization team to design and implement a web-based analysis dashboard using Flask and Python REST APIs, streamlining performance metric analysis and increasing the decision-making efficiency.
2023 — 2023
San Mateo, California, United States
Contribute effort to the SQL team, which is responsible for core query processing capabilities and leading performances.
Automated the monitoring process of SQL quality assurance, which saved 159 hours/year of each engineer resource.
Investigated query minimizer and integrated it into the workflow that increase the monitoring efficiency by up to 69%.
Created design docs to capture technical requirements and convey complex concepts in a clear and concise manner.
Utilized algorithms (e.g., divide-and-conquer) and Object-Oriented programming flexibly to solve real-world problems.
2022 — 2022
Bellevue, Washington, United States
1. Built solutions that continuously validate political preferences of 1.96 billion Meta users are not used in Ads targeting.
2. Helped Core Ads team solve the challenge of not supporting a fully self-serving onboarding service and automated the whole process by developing a test platform with React and JavaScript Relay framework to fetch and manage GraphQL data. 10+ product teams (e.g., Facebook, Instagram, and Messenger) benefit from it.
3. Utilized distributed SQL query engine Presto to query trillions of entries from Hive, a data warehouse based on Hadoop, and feed the data into the dashboard which increased the decision-making efficiency for the product teams.
4. Applied data structure Graph to define the relationship between users and events and used algorithms like Depth First Search (DFS) to recursively parse the cascaded information from UI to generate the corresponding Presto query.
5. Implemented CRUD with GraphQL queries over relational database XDB and operated on Petabytes of data.
2019 — 2020
Beijing, China
Develop an anomaly detection system based on metric data (e.g. mem_used, cpu_used, vm_pgfault) from thousands of cloud servers to ensure Elastic Compute Service;
Launched a real-time root cause analysis algorithm;
Performed data cleaning, aggregation, and analysis on up to 1.5M time-series data points;
Designed an automatic thresholds selection mechanism of anomaly alarming;
Deployed network microservices for QA to detect anomaly and locate root causes on the data platform.
Education
University of Southern California
Master of Science - MS
Nankai University