大模型算法工程师 - 生活服务
Core
Building an AI intelligent customer service system for lead advertising scenarios to improve customer acquisition efficiency and lead quality through high-conversion dialogue and intent recognition.
Role type
Senior Large Model Algorithm Engineer (Intelligent Customer Service)
Builds
AI intelligent customer service agents, dialogue systems, and reward models for lead conversion.
Domain
Advertising technology and local life services
Deliverable
production ML models
Required skills
Large language model (LLM) training and fine-tuning, Reinforcement Learning from Human Feedback (RLHF), Reward Model (RM) training, Agent Harness architecture, Multi-turn dialogue management, Intent recognition, Python, C++, Java
Preferred skills
RAG, Prompt engineering, Agentic RL, OPD (Online Policy Distillation), Multi-modal large models, Ad algorithm background (CTR/CVR)
Technologies
LLM, RAG, Agent, RLHF, DPO, GRPO, OPD, Python, C++, Java
Responsibilities
Design and build the algorithm system for intelligent customer service leads; Develop core algorithms for intent recognition, multi-turn dialogue, and script generation; Architect and evolve the Agent Harness environment for stability and iteration; Build and iterate reward models and alignment systems (RLHF/DPO/GRPO); Explore AIGC landing pages and multi-modal customer service applications; Collaborate with product, engineering, and operations to drive business results.
Seniority
Senior, hands-on IC
