Copilot Studio Architect

Copilot Studio Architect

Location: Pan India

Role Summary

We are looking for an Agentic AI Architect to own how custom agents are designed, governed, and scaled across our clients' Microsoft ecosystems. This is a design-authority role centred on Copilot Studio, but it reaches into the wider agent landscape: M365 Copilot extensibility, Power Platform, Azure AI, and the governance that keeps all of it safe. You set the architecture and standards, decide when an agent is the right answer and when it is not, and ensure what the teams build is secure, grounded, cost-aware, and maintainable. You stay technical enough to prototype and challenge a design, but your primary output is direction, standards, and governance.

Key Responsibilities

• Own the end-to-end architecture for agentic AI solutions, from use-case qualification through to a documented target design.
• Make the key platform decisions: Copilot Studio versus M365 Copilot agents versus pro-code (Azure AI Foundry / SDK), low-code versus pro-code orchestration, and where each is justified.
• Define grounding and knowledge architecture: which sources, how they are indexed and secured, and how retrieval quality is maintained over time.
• Design the security and identity model for agents, including Entra ID, least-privilege access, connector governance, and data boundaries.
• Establish governance for agentic AI: DLP policies, environment strategy, ALM, responsible AI guardrails, monitoring, and cost management for consumption/messaging.
• Provide architectural governance and design review over the agents that senior developers and teams deliver.
• Advise clients and internal stakeholders on agentic AI roadmaps, feasibility, and effort/complexity trade-offs, and support proposals and SOWs with credible designs.
• Anticipate risks around accuracy, data exposure, licensing/consumption cost, and long-term supportability, and design around them.

Must-Have Skills

• Architecture-level command of Copilot Studio and the broader Microsoft agent ecosystem (M365 Copilot extensibility, agents, connectors).
• Deep Power Platform expertise including environment strategy, ALM, and DLP/governance.
• Strong grasp of grounding, retrieval, prompt orchestration, and the practical limits of generative AI.
• Solid Entra ID, Microsoft Graph, and M365 security/identity knowledge.
• Ability to produce clear architecture artifacts: solution designs, data-flow and grounding diagrams, and governance frameworks.
• Understanding of responsible AI, data governance, and consumption/cost models for agent workloads.

Good-to-Have Skills

• Ability to reason about RAG at the design level: retrieval quality, chunking strategy, and knowledge-source architecture across the estate.
• Azure AI Foundry, Azure OpenAI, and pro-code agent frameworks (Semantic Kernel, LangChain) for hybrid and fully pro-code solutions.
• Python for pro-code agent and integration scenarios where low-code is not sufficient.
• Azure integration services (Functions, Logic Apps, API Management) for agents that reach beyond M365.
• SPFx / Power Apps for custom agent surfaces and embedding.
• SharePoint Online and on-premises (2019/SPSE) knowledge for enterprise content grounding and hybrid scenarios.
• PowerShell for M365 administration and automation.

Experience & Qualifications

Experience: Typically 10+ years across M365 / Power Platform, with demonstrable architecture ownership and real agentic AI or Copilot delivery.

Certifications: Architect-level certifications (PL-600, Azure/M365 architect credentials, and Copilot/agent-focused ones) are strongly preferred.

Soft Skills

• Able to hold a design position with clients and set realistic expectations on what agents can safely deliver.
• Strong written communication for architecture documentation and proposal support.