Research Engineer, Agentic Video Editing - TikTok

We are building the next generation of AI-native creation tools powered by intelligent AI agents. Our mission is to transform how people create by embedding agentic AI directly into creative workflows—from AI-powered camera effect generation to AI-driven video editing and storytelling.

Our team operates at the intersection of AI research, systems engineering, and creative product development, translating advances in agentic AI into scalable, real-world creation tools used by millions of creators. We work closely with product and design partners to ensure that sophisticated AI systems result in simple, powerful, and intuitive creative experiences.

Responsibilities - Build end-to-end agentic video editing systems, working across the agent harness, backend services, and evaluation pipelines. Own scoped features and improvements from prototyping through production. - Develop and improve the agent harness, including planning, tool use, context management, and error recovery for multimodal editing workflows. - Build reliable backend services for agent execution, session management, and media processing, and improve system latency, cost, and reliability. - Develop evaluation and feedback loops, including LLM judges, benchmarks, and human review workflows, to measure editing quality, diagnose failures, and catch regressions. - Propose new ideas and adapt advances from research and industry into practical experiments, using evaluation results and user feedback to continuously improve agent capabilities and video quality. - Collaborate with research, engineering, and product partners to turn successful experiments into useful editing experiences for creators.

Minimum Qualifications - Bachelor’s or Master’s degree in Computer Science, AI, or a related field, or equivalent practical experience. - Strong programming skills in Python and solid software engineering fundamentals. - Foundational knowledge of machine learning and large language models. - Hands-on experience building an LLM application, AI Agent, or multimodal AI project through coursework, research, internships, or personal projects. - Ability to design experiments, analyze failures, and turn findings into working improvements. - Interest in video creation and building useful tools for creators.

Preferred Qualifications - Experience with tool-use agents, context management, or debugging and improving multi-step agent workflows. - Familiarity with video understanding, audio and visual processing. - Experience with evaluation datasets, human annotation, or model-based evaluation. - Familiarity with video editing tools or media processing libraries such as FFmpeg. - Relevant research, open-source contributions, or projects that demonstrate the ability to build and evaluate AI systems.