Agent策略工程师（J104988）
Core
Design and iterate strategies for health content generation agents, building automated inspection and self-evolution mechanisms based on Agentic Looping to drive continuous quality improvement.
Role type
Senior IC agent strategy engineer (health domain)
Builds
Automated health content generation agents with self-evolving capabilities and multi-tool orchestration
Domain
Healthcare + Large Language Models (LLMs) + Agentic Systems
Deliverable
production ML models
Required skills
Complex system design, LLM fine-tuning and alignment (SFT/DPO/RLHF), Agentic Looping architecture, Harness engineering, Data engineering, Model evaluation
Preferred skills
Health/medical domain knowledge, Information Retrieval (IR), Knowledge Graphs, Large-scale inference systems
Technologies
Agentic Looping, SFT, DPO, RLHF, Harness engineering
Responsibilities
Design and iterate agent strategies for health content generation; Optimize task scheduling and intelligent routing for core tools (doctor/drug search); Build harness engineering systems to quantify model/strategy impact; Lead rapid experimentation and scaling of new LLM/agent technologies in health scenarios
