Responsibilities
1.Architect backend MCP services to enable seamless context exchange between browsers and LLMs.
2.Implement secure multi-tenant gateways with dynamic API key routing, request-level caching, and adaptive throughput control.
3.Develop a unified tool invocation layer exposing JSON-RPC interfaces and integrating with Chromium extensions.
4.Collaborate with frontend and browser teams to ensure ultra-low-latency MCP signaling(<100 ms) and support 10K+ QPS scalability.
5.Evolve MCP specifications aligned with open-source developments and maintain backward compatibility.
Requirements
1.3+ years of experience delivering Python backend systems, proficient in asyncio, FastAPI, pydantic, and high-performance libraries.
2.Strong background in distributed caching (e.g., Redis Streams), messaging systems(Kafka, Pulsar), and PostgreSQL tuning.
3.Hands-on expertise in authorization/authentication protocols (OAuth 2.0, JWT, SAML) and multi-tenant isolation strategies.
4.Skilled in container orchestration and CI/CD practices using Docker and Kubernetes.
5.Demonstrated ability to interpret and implement English-language technical documentation and community contributions.
Preferred Qualifications
1.Contributions to MCP or related official repositories.
2.Experience with Chromium Extension / Native Messaging, WebSocket, and Chrome DevTools Protocol.
3.Experience building Agent backend systems (e.g., AutoGPT, Copilot, ChatGPT Plugin).
4.Familiarity with LLM inference acceleration (vLLM, TensorRT-LLM) or prompt caching optimization.