AIGC&大模型评测专家 - TikTok研发
Core
Build and optimize the full-link evaluation system for TikTok's AIGC and large model products, covering dataset construction, metric systems, self-developed evaluation models, automation engineering, and Agent workflows.
Role type
Senior IC AIGC and Large Model Evaluation Engineer
Builds
Automated evaluation pipelines, multi-dimensional evaluation datasets, and self-developed evaluation models for AIGC products
Domain
AI/ML, Large Language Models (LLM), AIGC
Deliverable
production ML models
Required skills
LLM principles, Transformer architecture, training processes, MLOps/LLMOps, platform product construction, automated pipeline design, data-driven root cause analysis, cross-team collaboration
Preferred skills
0-to-1 evaluation infrastructure build experience, architecture perspective, prompt strategy optimization
Technologies
MLOps, LLMOps, Agent Workflow, automated data production tools
Responsibilities
Design and develop evaluation tools and automated pipelines; construct layered multi-dimensional evaluation datasets; build a unified evaluation metric system integrating model, engineering, and product effects; perform root cause analysis and drive optimization from algorithm, data, and engineering dimensions; track AIGC industry trends and output analysis reports for technical decision-making.
