Responsibilities:

• Engineer Core Components: Lead the development of the orchestration service and Agent-to-Agent (A2A) bus, writing high-performance Python code that routes thousands of requests per minute.
• Build Complex Agents: Implement "Horizon 3/4" autonomous agents (e.g., self-correcting coding agents, complex supply chain optimizers) using LangGraph or AutoGen hosted on AKS.
• Integrate Data & AI: Build robust integrations between Databricks (Unity Catalog) and our agent runtime, ensuring agents have access to real-time, governed business data.
• Optimize Performance: Tackle the hard problems of agentic AI—latency reduction, token cost optimization, and managing distributed state across long-running workflows.
• Mentor & Review: Conduct code reviews for junior engineers and divisional teams, ensuring that all agent code adheres to our Zero-Trust security practices and software engineering standards.
• Develop Tools (MCP): Build and maintain a library of Model Context Protocol (MCP) compliant tools that allow agents to interact with SAP, ServiceNow, and other enterprise APIs.
  
  

Qualifications:

• Engineering Mastery: You are a software engineer first. You write clean, testable, production-grade Python (FastAPI, Pydantic) and are comfortable with async/await patterns and concurrency.
• Container Native: You don't just run Docker locally; you have experience deploying and debugging complex applications on Kubernetes (AKS).
• Agentic Frameworks: Deep, hands-on experience with LangChain, LangGraph, or Semantic Kernel. You know how to debug a "looping" agent and how to manage conversation memory effectively.
• Azure & Databricks: Practical experience with Azure AI Foundry for model serving and Databricks for data engineering.
• Problem Solver: You enjoy debugging non-deterministic systems and finding creative ways to make LLMs reliable (e.g., retry logic, output parsing, verification steps).

Years of experience:

• 5