Agentic Architect
Core
Design and architect agentic software systems, including multi-agent applications and development harnesses, to connect business needs with technology strategy.
Role type
Senior Agentic Architect (LLM & Multi-Agent Systems)
Builds
Governed, auditable software factories and multi-agent applications on enterprise stacks.
Domain
Enterprise IT, Generative AI, Agentic Workflows
Deliverable
production ML models | product features
Required skills
Software architecture, LLM system design, Coding agents (Claude Code, Codex CLI), LangGraph, LangChain, Python, MCP (Model Context Protocol), LLM evaluation (LangSmith, Langfuse), API integration (Anthropic, OpenAI, Azure, Vertex), Tool-use patterns, Enterprise cloud (Azure/GCP), Kubernetes/OpenShift, CI/CD.
Preferred skills
Platform engineering, GitHub Actions, Azure DevOps, Vector stores, Retrieval, Regulated environments (banking, healthcare).
Technologies
LangGraph, LangChain, LangSmith, Langfuse, Claude Code, Codex CLI, MCP, Kubernetes, OpenShift, Azure, GCP, Anthropic, OpenAI, Azure OpenAI, Vertex AI.
Responsibilities
Design agentic development harnesses (hooks, skills, sub-agents, permissions, context, quality gates, traceability); Define intent-to-spec-to-plan-to-diff chains with controls; Architect multi-agent systems with state management and orchestration; Implement evaluation strategies and regression testing; Ensure observability, cost management, and security (prompt injection, guardrails); Apply architectures to legacy code migration; Create reference architectures and ADRs.
Seniority
Senior, hands-on IC
