大模型安全攻防工程师-抖音研发
Core
Build LLM for Security architecture and solutions for Douyin's security blue team, focusing on vulnerability detection, code analysis, and privacy compliance testing.
Role type
Senior IC LLM security engineer (offensive/defensive)
Builds
Security evaluation frameworks for LLM lifecycle (training, deployment, application) and content/Agent security
Domain
AI/ML security, Large Language Models
Deliverable
production ML models
Required skills
LLM deployment and optimization, adversarial attack research (jailbreak, model stealing, adversarial examples, data poisoning), prompt engineering, RAG design, vulnerability detection, code analysis
Preferred skills
Go, Python, Java, C++
Technologies
LLM frameworks, security testing tools
Responsibilities
Design and implement LLM security solutions, conduct security assessments across the LLM lifecycle, track and apply global LLM security research and attack/defense techniques
Seniority
Senior, hands-on IC
