保险科技-大模型高级工程师
Core
Design and build post-training pipelines for large language models (LLMs) to enhance professional Q&A, complex reasoning, and compliance risk control in insurance scenarios.
Role type
Senior IC machine-learning engineer (insurance domain)
Builds
Production LLMs for insurance underwriting, claims, and health consultation
Domain
Insurance + Large Language Models
Deliverable
production ML models
Required skills
LLM post-training (SFT, RLHF, DPO, GRPO), Reward Model development, dataset construction (instruction/preference), data cleaning and evaluation, PyTorch, Transformers, DeepSpeed, Megatron-LM, vLLM, Agentic-RL, RAG, knowledge graphs
Preferred skills
Insurance/financial domain knowledge, API integration for business simulation, automated evaluation systems
Responsibilities
Develop reward models and align LLMs for insurance-specific tasks; Construct and curate high-quality instruction datasets for underwriting and claims; Build Agentic-RL simulation environments for tool calling; Establish automated evaluation frameworks for model performance tracking.