Senior Software Engineer

Jane Sheng, MS

Sunnyvale, CA ***** 352-***-**** ****************@*****.*** Linkedin

Experienced software engineer with in-depth knowledge and hands on experiences in: Technical Leadership Data Analytics Data Processing Collaboration and Communication Parallel and Distributed Systems Project Management Machine Learning Code Review and Quality Assurance Mentoring and Knowledge Sharing Integrity and Security A/B Testing

TECHNICAL SKILLS

Languages: Python Hack (PHP) Java JavaScript SQL Technologies: GraphQL React Relay REST APIs MySQL PostgreSQL Apache Hive ZippyDB Presto PROFESSIONAL EXPERIENCE

META, Menlo Park, CA May 2019 - April 2025

Senior Software Engineer

• GNN Recidivism Auto Enforcement: Led four engineers and collaborate with 10+ XFN teams to built machine learning and software solutions to detect and mitigate recidivism risk on the Facebook developer platform, which supported millions of third-party apps and billions of users. Enhanced system scalability by over 10 and boosted data labeling quality by 40% through data-driven insights. Designed and built robust data processing infrastructure and pipelines to train a GNN model with precision above 90%. Implemented infra to automatically escalate recidivism apps and perform enforcement actions at scale immediately. It prevented more than 50k recidivist apps from abusing the platform.

• LLM Categorization: Led two engineers to built a novel integrity risk identify process by automatically categorizing reactive user reports with Llama. It helped the team in making priority decisions of two projects. It will continue to be used by both my team and XFN teams in future roadmapping.

• Significant Binary Change Detection: Collaborated with three XFN teams to develop engineering solutions to detect and prevent the binary updates that presented high risk of integrity, compliance, and security in Horizon Store.

• Open Store Captcha: Quickly ramped up on required front end technologies to implement a reliable captcha feature for preventing scripted app submission spam risk in Horizon Store Connect. It proved the hypothesis about scripting problem in store and protected the app submission review process.

• Policy Tagging System: Collaborated with 3+ XFN teams and launched the tagging system, which aimed at filtering out irrelevant data from all enforcement flows for audit reporting. It saved audit data preparing time from several weeks to one day.

• Intelligent Launch Decision Platform: Partnered with data scientists to conduct in-depth data analysis and identified potential opportunities related to the problem. Led two engineers to build solutions and software infra to improve the launch quality with higher guidance adherence. Besides, the launch review efficiency was largely improved by increasing guideline automation rate from 27% to 70%.

PDF SOLUTIONS, San Jose, CA September 2012 - March 2019 CV Design Engineer

• Developed programmable design tools, using Ruby to generate logical circuit cell automatically.

• Designed digital circuits and generated cells using the developed tools to help improve wafer yield. EDUCATION

Master of Science (MS) in Electrical and Computer Engineering UNIVERSITY OF FLORIDA, Gainesville, FL

Bachelor of Science (BS) in System and Information Engineering SOUTHEAST UNIVERSITY, Nanjing, China

Contact this candidate