Senior Research Scientist, TikTok - Trust and Safety
Core
Design, train, and deploy foundation models (LLM/VLM) for content safety, risk detection, and moderation automation on TikTok's global platform.
Role type
Senior Research Scientist (Multimodal Foundation Models)
Builds
General-purpose foundation models and multimodal safety annotation frameworks for risk detection and compliance.
Domain
Internet safety, Trust and Safety, Multimodal AI
Deliverable
production ML models
Required skills
Large model development, NLP, Computer Vision, RLHF, multimodal learning, scalable architecture design, training pipeline optimization
Preferred skills
Publications in top AI conferences (NeurIPS, ICML, ICLR, CVPR, ICCV), experience in competitive AI challenges
Technologies
LLMs, VLMs, GPU clusters, RLHF frameworks
Responsibilities
Lead design and deployment of foundation models for content safety; enhance model generalization in multilingual/low-resource scenarios; design multimodal annotation frameworks; optimize large-scale training pipelines; integrate safety models into moderation workflows
Seniority
Senior, hands-on IC
