Agent Research and Development Engineer
Beijing-based AI company that develops large language models and AI-native products, including hosted model APIs and the Baixiaoying assistant.
Funding history
About Baichuan AI (Baichuan Intelligent Technology)
Baichuan Intelligent Technology builds general-purpose and medical large-language models, provides developer-facing API services, and operates AI-native applications.
Skills
Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will define cognitive architectures that connect perception, planning, decision-making, and reflection for AI agents. You will build interfaces between agents and the real world, enabling models to perceive environments, call tools, and execute actions. You will develop reliable knowledge pipelines covering retrieval, reasoning, generation, and validation. You will build long-context memory capabilities and turn advanced agent capabilities into highly available, observable, and evolvable engineering foundations.
Requirements
- Bachelor's degree or above in computer science, artificial intelligence, or a related field
- At least 3 years of backend or systems development experience
- Proficiency in C++ or Go
- Strong engineering and system design skills
- Deep understanding of distributed systems, high-concurrency services, and microservice architecture
- Understanding of large language model limitations and failure modes
- AI agent core-capability experience is preferred
- Experience with retrieval systems, memory systems, or reasoning optimization is preferred
- AI implementation experience in a professional vertical such as healthcare, research, or finance is preferred
Responsibilities
- Define AI agent cognitive architectures for perception, planning, decision-making, and reflection
- Build interfaces that enable agents to perceive environments, call tools, and take actions
- Develop knowledge pipelines for retrieval, reasoning, generation, and validation
- Build multi-turn and long-context memory capabilities
- Turn advanced agent capabilities into highly available, observable, and evolvable engineering infrastructure