Drimva

TikTok

Machine Learning Engineer Graduate (Ads Signal & Measurement) - 2027 Start

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

Education
Completing or recently completed a Bachelor's or above degree in a listed or related technical discipline.

About the job

Build machine-learning models and large-scale systems for TikTok Ads signal quality, identity resolution, attribution, and measurement, and apply their outputs to ad ranking and optimization.

What you'll do

  • Develop signal-quality models, identity-resolution methods, and attribution and incrementality measurement.
  • Feed signal, identity, and attribution outputs into ads ranking, and help build experimentation infrastructure and data pipelines.
  • Explore LLM-powered understanding and correction of advertiser conversion signals; collaborate to deploy ideas in production.

What they ask for

  • Completing or recently completed a Bachelor's or higher degree in Computer Science, Statistics, Mathematics, Electrical Engineering, or a related technical discipline.
  • Foundation in machine learning and statistics, including hypothesis testing, regression, and probabilistic models; causal inference is ideal.
  • Strong programming skills in at least one of Python, Go, Java, or C/C++, and understanding of data structures and algorithms.
  • Hands-on ML experience through coursework, research, internships, or personal projects.

Nice to have

  • Experience in computational advertising, ranking, causal inference or experimentation, entity resolution or graph learning, anomaly detection, or applied LLMs/NLP.
  • Publications or strong competition performance; experience with Spark, Flink, PyTorch, or TensorFlow.

Tools

Python, Go, Java, C/C++, Spark, Flink, PyTorch, TensorFlow

AI focus

Machine learning for signal quality, identity resolution, attribution, and measurement; apply large language models to understand, classify, assess, and automatically correct advertiser conversion signals.

Model-written summary · the original posting is authoritative

First seen Oct 28, 2026 · Last verified Nov 6, 2026 (UTC)