AI Agent Architect

About the Role We’re seeking an exceptional AI Agent Engineer/Architect to design and implement advanced, protocol-compliant agent systems. The ideal candidate will have deep hands-on experience building Model Context Protocol (MCP) and Agent-to-Agent (A2A) integrations, as well as a background in data catalog engineering within major hyperscaler ecosystems. This is a high-impact, hands-on role focused on building next-generation AI agent interoperability across enterprise data platforms. 🔧 Key Responsibilities • Architect and develop AI agent frameworks compliant with MCP (Model Context Protocol) and A2A (Agent-to-Agent) communication standards. • Design integrations between AI agents and Microsoft Fabric, OneLake, and Purview data catalogs. • Build high-performance data access, indexing, and semantic enrichment pipelines for metadata and catalog intelligence. • Develop APIs, connectors, and microservices using Rust, C#, and Python. • Optimize agent context management and secure cross-agent communication. • Work closely with AI, data, and platform engineering teams to ensure scalability, interoperability, and compliance. • Contribute to the architecture of a next-gen data intelligence product leveraging LLMs, metadata-driven reasoning, and context-aware agents. 🎯 Required Skills & Qualifications • Prior experience as a Systems Engineer, Architect, or Senior Developer in Dataplex, Databricks Unity Catalog, or OneLake/Purview teams. • Strong programming expertise in Rust, C#, and Python. • Deep understanding of Model Context Protocol (MCP) and Agent-to-Agent (A2A) interoperability standards. • Experience with data catalog systems, metadata services, and enterprise data governance. • Familiarity with Microsoft Fabric ecosystem and OneLake architecture. • Proven ability to build distributed, secure, and high-performance backend systems. • Excellent problem-solving, system design, and communication skills. 🧩 Nice to Have • Familiarity with LLM orchestration frameworks (LangChain, Semantic Kernel, AutoGen). • Background in cloud-native development (Azure, GCP, or AWS). • Knowledge of metadata graph systems or knowledge catalogs.