Drimva

TikTok

Machine Learning Engineer, Ecommerce Risk Control

San Jose, California · Full-time · Posted Oct 28, 2026

Experience
3+ years of hands-on experience building and delivering machine-learning models for large-scale projects; preferred: 4+ years of industry experience in predictive modeling and analysis.
Education
Master’s degree in Computer Science, Mathematics, Machine Learning, or another relevant STEM field (finance may be relevant for financial fraud roles).

About the job

Build and deploy machine-learning solutions, data infrastructure, and risk metrics to detect and mitigate threats across TikTok e-commerce platforms.

What you'll do

  • Invent, test, and deploy machine-learning algorithms and prototypes for e-commerce risk control.
  • Build data-querying infrastructure and monitor risk metrics to uncover attacks and improve defenses.
  • Collaborate with cross-functional teams and communicate findings, recommendations, and technical results.

What they ask for

  • Master’s degree in Computer Science, Mathematics, Machine Learning, or another relevant STEM field (finance may be relevant for financial fraud roles).
  • 3+ years of hands-on experience building and delivering machine-learning models for large-scale projects.
  • Programming experience in Java, C++, Python, or a related language.
  • Experience developing models and visualizations using programming or scripting languages, and using forecasting, machine-learning, and statistical tools.

Nice to have

  • PhD in CS, Machine Learning, Statistics, Operations Research, or a relevant field; 4+ years of industry experience in predictive modeling and analysis.
  • Experience with product, operations, and engineering teams, or in e-commerce/online fraud or risk control.

Tools

Java, C++, Python, Scala, R, Ruby, Matlab

AI focus

Develop and deploy machine-learning models to prevent and detect e-commerce risks, including account integrity threats, incentive abuse, and malicious activity.

Model-written summary · the original posting is authoritative

First seen Oct 28, 2026 · Last verified Nov 4, 2026 (UTC)