Applied Scientist, Multimodal Safety & Moderation

TikTok; Seattle, Northern (WA, KY); Hybrid; USD 207,000 - 368,000; Full time

Benefits offered by this job

Health insurance

401(k) with company match

Paid parental leave

Disability coverage

Job summary

TikTok in Seattle is seeking an Applied Scientist for Trust and Safety (Multimodal Foundation Model). You will work on state‑of‑the‑art moderation systems, focusing on large‑scale multimodal models, RL‑driven decision making, and robust safety tooling.

The role targets recent PhDs ready to ship impactful research into production. The project emphasizes cross‑modal alignment, agentic moderation, and safety improvements across 200+ languages and content types.

Qualifications

• PhD completed or near completion in CS/AI or related field.
• Strong understanding of LLM research and practical implementation.
• Proficiency in Python, Rust or C++ and DL frameworks.
• Experience with distributed computing and RL/SE for AI.

Responsibilities

• Develop and deploy multimodal safety foundation models for moderation.
• Research agentic moderation systems with RL and tool integration.
• Advance MoE architectures, cross‑modal alignment, and robust evaluation.
• Collaborate with cross‑functional teams to ship scalable safety solutions.

Skills

Python

Rust

C++

Multimodal ML

RL

LLM research

Education

PhD in CS/AI

Tools

PyTorch

DeepSpeed

Megatron-LM

Job description

TikTok in Seattle is seeking an Applied Scientist for Trust and Safety (Multimodal Foundation Model). You will work on state‑of‑the‑art moderation systems, focusing on large‑scale multimodal models, RL‑driven decision making, and robust safety tooling.

The role targets recent PhDs ready to ship impactful research into production. The project emphasizes cross‑modal alignment, agentic moderation, and safety improvements across 200+ languages and content types.