Description

Design and build production-grade multi-agent AI systems using Azure AI services, optimizing for latency, cost, and reliability.

This role is on-site.

Responsibilities

• Architect multi-agent systems using Azure AI Agent Service and Microsoft Agentic Framework, defining roles, memory, and orchestration patterns.
• Implement MCP (Model Context Protocol) services with FastMCP, enabling communication via SSE and structured tool contracts.
• Build advanced RAG pipelines using Azure AI Search, handling document ingestion, embedding strategies, and context window optimization.
• Orchestrate multi-step reasoning workflows with fault tolerance, retry mechanisms, and state management across distributed systems.
• Design evaluation frameworks and implement deep observability using Azure AI Foundry, Azure Monitor, and custom audit hooks.

Required Skills

• 6+ years in software engineering with strong expertise in Python.
• 2+ years of hands-on experience building production-grade LLM/AI applications.
• Experience integrating with .NET-based systems, APIs, and microservices.
• Proficiency with Azure OpenAI, Azure AI Foundry, and Azure AI Agent Service.
• Knowledge of RAG architectures, prompt engineering, tokenization, and context windows.
• Experience with distributed data systems: Cosmos DB, Redis, Azure Blob Storage, and AWS (EC2, S3, RDS).
• Understanding of event-driven systems and scalable agent-tool interaction layers.

Preferred Skills

• Experience with multi-agent systems, tool-using agents, and MCP or FastMCP frameworks.
• Basic to intermediate machine learning experience (Scikit-learn, XGBoost) including classification, regression, and clustering.

Key Skills

Education

Bachelor's degree