Research Engineer, Agentic Video Editing - TikTok
Core
Building end-to-end agentic video editing systems that integrate AI agents into creative workflows for video creation and storytelling.
Role type
Senior IC research engineer (agentic AI systems)
Builds
Scalable, real-world AI-native creation tools for video editing and storytelling
Domain
AI research, systems engineering, and creative product development
Deliverable
production ML models
Required skills
Python, software engineering fundamentals, machine learning, large language models, LLM application development, AI agent development, multimodal AI project experience, experiment design, failure analysis, system improvement
Preferred skills
tool-use agents, context management, multi-step agent workflow debugging, video understanding, audio and visual processing, evaluation datasets, human annotation, model-based evaluation, video editing tools, media processing libraries (FFmpeg), open-source contributions
Technologies
Python, LLMs, FFmpeg
Responsibilities
Build end-to-end agentic video editing systems across agent harness, backend services, and evaluation pipelines; Develop and improve the agent harness for planning, tool use, context management, and error recovery; Build reliable backend services for agent execution, session management, and media processing; Develop evaluation and feedback loops including LLM judges and benchmarks; Propose new ideas and adapt research advances into practical experiments; Collaborate with research, engineering, and product partners to turn experiments into useful editing experiences