可信隐私计算高级研究员-安全与风控
Core
Design and implement innovative cryptographic protocols and trusted solutions for privacy-preserving data flow in cross-organizational large language model (LLM) scenarios.
Role type
Senior Research Scientist (Privacy Computing & LLM Security)
Builds
High-concurrency, large-scale privacy computing systems enabling 'data available but invisible' for business growth.
Domain
AI/LLM Security, Cryptography, Distributed Systems
Deliverable
production ML models
Required skills
Multi-party computation, Federated learning, Homomorphic encryption, Differential privacy, Trusted Execution Environments (TEE), Large model architecture (Transformer), Fine-tuning (LoRA/PEFT), Distributed computing (Spark/Hadoop), Linux, C/C++/Python/Go
Preferred skills
LLM privacy protection实战 experience, Leading open-source privacy computing projects, Publishing high-quality papers at top security conferences
Responsibilities
Explore and deploy new applications of privacy computing in cross-domain LLM scenarios; Design and implement innovative cryptographic protocols for LLM-specific security challenges (training data protection, model weight protection, prompt privacy, RAG isolation); Lead core architecture design and R&D of the privacy computing platform to optimize performance, stability, and scalability.
