Title : Microsoft Copilot Studio Developer – AI Agent & MCP at Somerset, NJ - Onsite
job description
Key Responsibilities
Explore, design and develop AI agents using Microsoft Copilot Studio.
Build and integrate MCP servers/tools to securely connect agents with enterprise applications, APIs, and data sources.
Understand different AI/GenAI use cases and translate them into practical technical solutions.
Build POCs/prototypes to evaluate different approaches and recommend the right solution.
Develop agent workflows using Power Automate, connectors, APIs, and Microsoft 365 services.
Implement multi-agent and agent-to-tool integration patterns for enterprise use cases.
Integrate AI solutions with existing enterprise applications and data platforms.
Configure authentication, authorization, security, governance, and data-access controls.
Integrate Copilot agents with Microsoft 365, SharePoint, Teams, Dataverse, Azure, and external platforms.
Work with business, data engineering, platform, security, and application teams to take solutions from idea/requirements through development and deployment.
Troubleshoot agent orchestration, tool invocation, grounding, and integration issues.
Partner with business and architecture teams to convert use cases into production-ready AI solutions.
Required Skills
Strong hands-on experience with Microsoft Copilot Studio.
Experience developing and integrating solutions using Model Context Protocol (MCP).
Strong knowledge of Power Platform, particularly Power Automate and Dataverse.
Experience with REST APIs, JSON, OAuth 2.0, Microsoft Entra ID, and API-based integrations.
Development experience with Python/C#/JavaScript/TypeScript, or equivalent.
Understanding of Generative AI, Large Language Models, AI agents, grounding, tool calling, and Retrieval-Augmented Generation (RAG).
Understanding of enterprise security, governance, responsible AI, and data protection.
Preferred Skills
Familiarity with Azure Databricks, structured and unstructured data, APIs, and enterprise data sources
Understanding of agent observability, testing, evaluation, deployment, and lifecycle management
