Tom Fan
Head of Trust & Safety, Video @Newsbreak
Signup · Get unlimited contacts
WORK HISTORY
Head of Trust & Safety, Video @Newsbreak
Mountain View, CA, US
Functionally lead Trust & Safety as the most senior operator across two distinct products: a video-first platform and a news-first platform with integrated video features. Own the design and scaling of systems that protect platform integrity, quality operations, and ensure product compliance. Lead cross-functional strategy and execution across content governance, policy, moderation infrastructure, vendor operations, and product safety.First 90 days: Created foundational dashboards and cross functional education on Trust and Safety. Establish SOP and governance standards for internal and external partnersReduced platform Badness(leakage) from 14% to 2.3% Key contributor to platform reinstatement on the Apple Store following compliance risk mitigation.First 180 days:Designed and implemented policy delimiter system to improve model; recall 76%-> 92%, precision 48%-> 71%Improved vendor ROI by ~40% through operational and strategy changes; automation, removed low value tasks, upskilled team.
ABOUT TOM FAN
I love building things; I’m a systems architect at heart.Trust and Safety leader with 8 plus years building scalable moderation operations, policy programs, and trust systems across high volume consumer platforms. I lead cross functional initiatives spanning appeals and enforcement, content governance, vendor and BPO operations, QA, and safety tooling, translating policy and operational needs into product and engineering requirements.I operate proactively and with urgency. I identify where workflows break, where tools fall short, and where risk is emerging, then drive solutions from problem definition to rollout. I build dashboards and operating metrics, design SOPs and governance, and continuously optimize queues, escalation paths, and quality controls to improve both user outcomes and operational efficiency.I am fluent in balancing human review with model assisted decisioning. I partner with product, engineering, and operations teams to improve ML supported moderation and appeals by tightening policy logic, improving precision and recall, defining clear edge case handling, and ensuring manual review remains effective for nuance and high risk areas. My focus is measurable impact, fast iteration, and scaling safety without slowing the product.I work globally and cross culturally, and I am bilingual. I have partnered with teams and stakeholders across major global hubs and time zones, aligning on shared standards while adapting execution to local context.Outside of work, I tinker with ComphyUI, lift weights, MMA conditioning, binge Lord of the Rings Extended cut, hitch hike on bandwagons like OpenClaw, volunteer in an education based non-profit, and take too many pictures of my cats.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.