Design and deploy production-grade multi-agent workflows using advanced frameworks (e.g., LangGraph, CrewAI, AutoGen)
Implement complex orchestration patterns including intent detection, dynamic routing, state management, and hierarchical topologies
Build and govern Model Context Protocol (MCP) servers to expose enterprise tools, internal systems, APIs, and data sources as structured, auditable agent tool surfaces
Configure and scale AI Gateway platforms (LiteLLM, AWS Bedrock, Azure API Management, or Apigee AI Extensions) for multi-model routing, cost optimization, quota enforcement, and automated model fallback
Integrate robust guardrails directly into the model invocation layer
Own the design and end-to-end optimization of our RAG pipelines
Establish production LLM evaluation frameworks and implement red-teaming for prompt injection and adversarial inputs
Support and contribute to modern CI/CD pipelines tailored for agentic systems.
Requirements
6+ years of experience in enterprise backend engineering, microservices architecture, and cloud infrastructure (Docker, Kubernetes, CI/CD pipelines)
2+ years of hands-on experience building and deploying production-level LLM applications, with a strong focus on state machines, multi-agent frameworks, and advanced RAG
A strong understanding of least-privilege tool access, sandboxed execution, and full audit traceability within data-sensitive environments
Deep familiarity with Python or TypeScript, modern API Gateways, and AI observability tools.