大模型Agent研究工程师-Seed
Core
Research and engineering of Large Language Model (LLM) Agent systems, optimizing performance for complex task execution including planning, memory, and multi-agent collaboration.
Role type
Senior IC LLM Agent Research Engineer
Builds
Production-ready Agent systems for programming, research, operations, and knowledge work scenarios
Domain
Artificial Intelligence / Large Language Models / Multi-Agent Systems
Deliverable
production ML models
Required skills
LLM architecture, multi-modal models, agent system design, prompt engineering, tool use, task decomposition, context management, model orchestration, benchmark creation, automated evaluation pipelines
Preferred skills
Reinforcement Learning (RL), multi-agent systems, training optimization, RLHF
Technologies
LLM frameworks, multi-modal frameworks, orchestration tools, evaluation frameworks
Responsibilities
Design and implement benchmarks and evaluation metrics for large-scale agent tasks; Optimize key agent capabilities including harness, memory, and execution strategies; Collaborate with product and platform teams to drive agent adoption across various business scenarios.
