大模型数据工程师 - Seed Model
Core
Design and implement high-performance, scalable, distributed data infrastructure for large language models (LLMs) and multimodal AI, covering the full lifecycle from storage to analysis.
Role type
Senior IC data engineer (large model & multimodal)
Builds
Data infrastructure, multimodal training data pipelines, and data assets for LLMs and GenMedia products.
Domain
Artificial Intelligence / Large Language Models / Multimodal AI
Deliverable
production ML models
Required skills
C/C++, Python, Go, Ray, Spark, AI Coding tools, CLI, Agent frameworks
Preferred skills
Experience with data augmentation, data distillation, high-quality data filtering, industry-impactful LLM projects
Technologies
Ray, Spark, AI Coding tools
Responsibilities
Design and implement distributed data infrastructure for storage, collection, cleaning, and management; Develop data processing operators and pipelines for audio-visual multimodal training; Build a 'data-model-evaluation' loop using Agents to drive model iteration; Explore and implement innovative data strategies like data distillation and filtering.
Seniority
Senior, hands-on IC