MCP Protocol Engineer

Role overview Design and build infrastructure that lets AI models interact reliably with tools, codebases, cloud services, and business data through the Model Context Protocol (MCP). The work spans protocol design, secure execution environments, developer tooling, and coordination across distributed agent systems.

Responsibilities - Build and maintain performant MCP servers that connect models to external systems and data. - Create isolated environments for running agent tools safely, including WebAssembly or container-based sandboxes. - Extend protocols to support streaming, binary data, and communication across distributed MCP deployments. - Develop SDKs that help developers create and host MCP servers. - Improve schema validation, error handling, and documentation for the surrounding tools.

Requirements - Experience designing protocols or RPC systems, such as gRPC, JSON-RPC, server-sent events, or WebSockets. - Knowledge of type-safe schema validation approaches, such as JSON Schema or Protocol Buffers. - Proficiency in at least one of TypeScript, Go, or Rust. - Experience with secure sandboxing or isolated execution environments. - Care for developer experience, clear documentation, and robust error handling.

Nice to have - Contributions to MCP open-source projects or experience with AI agent frameworks. - Familiarity with identity, authentication, or capability-based authorization.

Benefits and work setup - Remote-first work with a globally distributed team. - High autonomy and end-to-end ownership, with an emphasis on outcomes. - Opportunity to contribute to open-source software used by developers.