Jonathan Wang Email & Phone Number

Fremont, Ca, Us

Fremont, Ca, Us

San Francisco, Ca, Us

- Managed two teams with a total of 9 engineers- Content Safety - Collaborated with Contributions, Services, and User Ops teams to protect users from unsafe and disruptive content- Traffic Quality - Worked closely with Ads teams to ensure that inorganic traffic is not billed - Protected systems by blocking bot and scraper traffic - Owned bot protection vendor selection end-to-end, coordinating Tech Sourcing, Legal, and Product - Introduced ML models to a rule-based bot filtering streaming pipeline

San Francisco, Ca, Us

- Created a new team of 4 engineers focused on protecting users from unsafe and disruptive content- Built team processes, charter, and roadmap for the new team- Collaborated with Contributions, Services, and User Ops teams to ensure content is protected

San Francisco, Ca, Us

- Managed a group of 3 engineers on the Spam team working on end-to-end machine learning projects to detect fake reviews- Ramped up another manager and several engineers to the team- Established processes to enable collaboration among the team working in multiple timezones / countries

San Francisco, Ca, Us

- Led projects to prototype, productionize, and deploy machine learning models to detect spam- Designed labeling tasks to gather exploratory and training data in a highly imbalanced class distribution- Worked closely with product management to evaluate spam policies, model performance, and business impact- Investigated a variety of infrastructure and model performance issues as part of the on-call rotation- Mentored multiple interns and new hires, scoping out ramp up projects and onboarding timelines- Spearheaded initiatives to improve team effectiveness by migrating to new frameworks and streamlining processes- Collaborated with other machine learning teams to exchange signals and centralize tooling

San Francisco, Ca, Us

- Worked on Spam team to train deep learning model to detect spammy images- Curated image data set and determined evaluation metrics- Deployed model into image upload pipeline to classify and handle incoming images

Berkeley, Ca, Us

Variable Selection for Palomar Transient Factory Data- Developed parallelized variable selection methods using Python and Apache Spark- Reduced model training time by 6x without loss in accuracy- Awarded 2nd place out of 62 posters in ACM student research competitionPartial Discharge Detection (in collaboration with Hyundai Electric & Energy Systems Co., Ltd.)- Identified a novel feature set and prediction models to classify partial discharge event signal data with 99% accuracy - Developed a filtering and threshold method to locate partial discharges in transformers up to 3x more accurately than existing methods (patent pending)

- Developed ASP.NET web applications to help automate protein production pipeline- Consolidated and restructured SQL databases- Integrated barcode printer into web application

Master Of Computer Science, Computer Science

Bachelor’S Degree, Computer Science, Statistics