Sr. AI Engineer - Microsoft Stack

Experience

Role Overview

1. Agentic Orchestration & Design

• Architect and develop complex multi-agent systems using the Microsoft Agent Framework (integrating patterns from Semantic Kernel and AutoGen).
• Implement Agent-to-Agent (A2A) communication and Model Context Protocol (MCP) to allow agents to discover and invoke enterprise tools dynamically.
• Design hybrid orchestration models: combining Agentic Reasoning (dynamic) with Workflow Orchestration(deterministic business logic).
• 2.
• AI Governance & Security (The Citadel Layer)
• Deploy and manage the Foundry Citadel Platform to ensure unified governance, cost control, and observability across all AI workloads.
• Implement AI Gateway patterns to enforce enterprise policies, content safety (Prompt Shields), and managed guardrails.
• Utilize Citadel’s Agent-Level Tracing and automated evaluations to monitor performance and compliance in real-time.

3. Pro-Code Backend & Data Engineering

• Build robust RAG (Retrieval-Augmented Generation) pipelines using Foundry IQ and Azure AI Search for high-accuracy, multi-hop reasoning.
• Develop backend services in Python or C# (.NET) to serve as the "brain" for AI agents, integrating with Microsoft Fabric and Dataverse.
• Automate the AI lifecycle (LLMOps) using GitHub Actions and Azure DevOps for CI/CD of prompt flows and model deployments.

4. UI & Frontend Integration

• Develop custom, responsive user interfaces using React or Streamlit that allow humans to interact with, supervise, and steer AI agents.
• Integrate agentic workflows into existing surfaces like Microsoft Teams, custom web portals, or the VS Code AI Toolkit.

Languages

• Advanced Python (FastAPI/Pydantic) and/or C# (.NET 8+).

Orchestration

• Microsoft Agent Framework, Semantic Kernel, or AutoGen.

Platform

• Azure AI Foundry (Studio/Hub), Azure OpenAI Service.

Governance

• Citadel (Foundry Citadel Platform), Microsoft Purview, or AI Content Safety is preffered.

Data/Search

• Foundry IQ, Azure AI Search (Vector/Hybrid), Cosmos DB.

Frontend

• React.js, Tailwind CSS, and experience with streaming LLM responses.

Tools

• VS Code AI Toolkit, GitHub Copilot, Prompt Flow.