Agent Network后端研发工程师-算力基础设施
Core
Design and develop the core system for Agent Network, building a unified access layer for Agent registration, routing, protocol adaptation, and session management.
Role type
Senior Backend Engineer (AI Infrastructure)
Builds
Unified Agent access layer, Model Gateway, MCP Gateway, and access control systems for external API/Tool calls.
Domain
AI Infrastructure / Agent Systems / Model Gateways
Deliverable
production ML models
Required skills
Go/C/C++, TCP/IP/HTTP/gRPC/WebSocket/SSE, network programming, system design for high concurrency/low latency, auth/authorization, rate limiting, circuit breaking, distributed systems.
Preferred skills
Envoy, LiteLLM, zero-trust gateway, permission systems, policy engines, AI platform development.
Responsibilities
Design and implement access control capabilities for Agent external API/Tool/service calls; build model gateway capabilities for multi-model routing and optimization; build MCP gateway capabilities for server/tool call management and governance; optimize model calling strategies based on cost, quality, and latency; follow up on frontier directions like LLM Gateway and Agent Runtime.
Seniority
Mid-Senior, hands-on IC