Agent应用架构工程师 - Seed Model
Core
Design and implement the architecture for core Agent scenarios in products like Doubao, enabling information retrieval, creation, and memory capabilities for conversational robots and deep AI applications.
Role type
Senior IC Agent Application Architect
Builds
Production-ready Agent systems for conversational robots, office AI applications, and general-purpose AI tools
Domain
Generative AI, Large Language Models (LLM), Multi-modal AI
Deliverable
production ML models
Required skills
LLM training and inference architecture, Agent system design, C++, Golang, Python, data structures, algorithms, system design, API development, code quality control
Preferred skills
Deep understanding of Agent concepts, strong problem-solving abilities, experience with long-context reliability and latency optimization
Technologies
C++, Golang, Python
Responsibilities
Design and implement Agent architecture for core product scenarios; Develop context engineering and tool design for office and general AI applications; Solve infrastructure challenges for large-scale Agent applications including task reliability and latency; Collaborate with model training teams to optimize training infrastructure and efficiency tools