Principal Software Engineer
Core
Design and ship MCP server/client infrastructure, agent-to-agent communication layers, and agentic workflow runtimes for enterprise AI orchestration.
Role type
Principal Software Engineer (AI Infrastructure & Agentic Systems)
Builds
MCP infrastructure, A2A communication layer, agentic workflow runtime, evaluation/observability tooling, and security controls for agent systems.
Domain
Enterprise AI, Agentic Orchestration, Multi-Agent Systems
Deliverable
production ML models | infrastructure
Required skills
MCP server/client design, A2A multi-agent communication patterns, Java/TypeScript, multi-step agentic workflow engineering, non-deterministic system observability, platform infrastructure architecture, security engineering for agent systems
Preferred skills
MCP/A2A spec contributions, open-source agent orchestration, agent security/adversarial testing, conference talks on agent orchestration
Technologies
Java, TypeScript, Node, MCP, A2A
Responsibilities
Design MCP infrastructure (servers, clients, schemas, auth, versioning), build A2A layer (discovery, delegation, handoff), develop agentic workflow runtime (planning, state management, human-in-the-loop), solve hardest stack problems end-to-end, engineer security (sandboxing, prompt-injection defenses), debug production failures in non-deterministic paths, make build-vs-buy decisions on agent infrastructure.
Seniority
Principal, hands-on IC
