Applied Scientist - Trust and Safety (Multimodal Foundation Model) - Global Frontier Tech Recruitment Program - 2027 Start (PhD)

TikTok
Seattle
Workplace: OnsiteFull timeFunction: Research & Scientific (R&D)Education: phdSkills: ["Problem-solving","Creative mindset","Research execution","Driving projects","Evaluation focus"]

Build multimodal AI models and agentic moderation systems to protect users and publishers at extreme scale. Work on MoE-based multimodal safety foundation models, RL-driven multi-step decision-making, and context/tool collaboration using flexible tool ecosystems. Improve training stability, routing optimization, inference acceleration, and robustness across 200+ languages while advancing evaluation of LLM and agent performance for generative content moderation.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
TikTok
TikTok
1 month ago

Applied Scientist - Trust and Safety (Multimodal Foundation Model) - Global Frontier Tech Recruitment Program - 2027 Start (PhD)

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 23 hours agoStatus: Live

Job Summary

Build multimodal AI models and agentic moderation systems to protect users and publishers at extreme scale. Work on MoE-based multimodal safety foundation models, RL-driven multi-step decision-making, and context/tool collaboration using flexible tool ecosystems. Improve training stability, routing optimization, inference acceleration, and robustness across 200+ languages while advancing evaluation of LLM and agent performance for generative content moderation.
Location: Seattle
Workplace: Onsite
Employment Type: Full time
Job Function: Research & Scientific (R&D)
Seniority: Graduate level

Key Responsibilities

  • •Develop multimodal moderation foundation models using large-scale sparse MoE, focusing on routing optimization, cross-modal alignment, and unified understanding/generation.
  • •Design agentic moderation systems that use reinforcement learning to improve multi-step decision-making and tool-call strategies.
  • •Implement context engineering and tool collaboration with dynamic context assembly and an MCP-based tool ecosystem.
  • •Improve training and inference performance through distributed computing optimization, verification, and inference acceleration.
  • •Evaluate AI systems, including LLM applications and agent development, for moderation scenarios and robustness.

Key Requirements

  • •Completing or having recently completed a PhD in Computer Science, Data Science, Artificial Intelligence, or a related field.
  • •Strong understanding of cutting-edge LLM research including long context, multimodality, alignment, and agent ecosystems, with practical implementation experience.
  • •Proficiency in Python, Rust, or C++ and experience working with deep learning frameworks such as PyTorch, DeepSpeed, Megatron, and vLLM.
  • •Strong understanding of distributed computing, performance tuning, and verification for training, finetuning, and inference.
  • •Familiarity with PEFT, RL, MoE, CoT, or LangChain is a plus.
Education:PhD / Doctorate
Skills:Problem-solvingCreative mindsetResearch executionDriving projectsEvaluation focus
Tech Stack:PythonRustC++PyTorchDeepSpeedMegatronVLLMPEFTRLMoECoTLangChainGRPOPPOMCPGraphRAGGPU

Company Brief

TikTok
Short-form video platform that lets users create, share, and discover entertainment content through algorithmic recommendations. It also offers advertising and creator tools for brands, influencers, and businesses.
Industry: Digital Media
Company Size: Enterprise (1,001+ employees)
Growth: Scaleup
Headquarters: Singapore, Singapore
Founded: 2016
WebsiteLinkedIn