Job description
Responsibilities
1. Responsible for the design and development of the core runtime links of the AI Agent, including task orchestration, model invocation, tool invocation, context management, and other modules.
2. Design and implement the security boundary system for the Agent: access control, sandbox isolation, multi-tenant resource and data isolation.
3. Build and maintain a tool/skill extension system that supports on-demand loading, third-party protocol integration, and plugin mechanisms.
4. Responsible for authentication and data isolation design in multi-user and multi-Agent scenarios.
5. Participate in the capability construction of the gateway layer: channel access, scheduled tasks, message queues, and other infrastructure.
6. Promote system security governance, identify and fix potential risks, and improve observability and fault diagnosis capabilities.
7. Participate in the design and evolution of the storage layer, balancing functional expansion and code maintainability.
8. Participate in the capability design of trading-related Agents, including tools and skill development in scenarios such as market data, account, and asset analysis.
9. Participate in the design of trading strategy Agents, transforming strategy logic (signal judgment, position management, risk control rules, etc.) into executable Agent capabilities and toolchains.
Requirements
Must-have
1. A full-time undergraduate degree or above, with more than 3 years of Golang backend development experience, familiar with Go concurrency model and standard library, capable of independently designing medium to large service modules.
2. Experience in distributed systems or multi-tenant backend, understanding access control models and configuration layering design.
3. Practical experience in containerized deployment and sandbox isolation.
4. Solid engineering judgment, able to identify gaps between design and implementation and drive improvements.
5. Good system design and documentation skills.
Bonus points
1. Experience in designing or developing trading strategy Agents, able to decompose trading strategies (coin selection/timing, position management, take profit/stop loss, risk control rules, etc.) into executable decision-making and tool invocation logic for Agents.
2. Understanding of the capability boundaries and risk control requirements of Agents in financial scenarios such as market data, accounts, and assets.
3. Experience in backend development of AI Agents / LLM applications, understanding basic concepts such as agent task loops and tool invocation.
4. Familiarity with design ideas for Agent systems such as access control, context compression, and sub-task orchestration.
5. Experience in code execution isolation or security sandbox.
6. Familiarity with database modeling, migration, and security practices for sensitive information storage.
7. Experience in frontend collaboration, able to assist in completing management backend integration.
8. Participation in security audits or issue rectification, able to identify common security risks.