[Hiring] Lead Power Platform Developer @Swan iT Recruitment

Role Description

This role sits at the point where the Microsoft 365 estate, the Power Platform and conversational AI meet. You will design and build business applications on Power Apps and Power Automate, and design and ship agents that sit on top of them in Copilot Studio.

We are looking for someone who thinks in terms of enterprise architecture rather than individual applications. The work involves:

• Deciding where a process belongs.
• Determining which parts of the Microsoft 365 and Azure stack should carry it.
• Understanding how data flows between them.
• Governance of the solution once it is live.

Building the solution is a large part of the job, but so is the reasoning that precedes it. Copilot Studio is a central skillset for this position, not a peripheral one. A meaningful share of current and pipeline work involves agents that act on enterprise data and trigger real business processes, and we need someone who has taken that kind of work into production rather than only through a proof of concept.

Qualifications

• Enterprise architecture.
• Microsoft 365.
• Microsoft Power Platform.
• Power Apps, both canvas and model-driven.
• Power Automate, including cloud flows and desktop flows.
• Copilot Studio, including agent design, knowledge grounding, actions and deployment.
• Dataverse data modelling and security.
• Power Fx.
• Application lifecycle management and CI/CD for the Power Platform.

Requirements

• Assess business processes and decide what belongs in the Power Platform, Microsoft 365, Azure, and what should not be built at all.
• Define environment strategy across development, test and production, including Dataverse provisioning, environment groups and managed environments.
• Design the data model, starting with Dataverse tables, relationships, business rules and calculated columns, and extending to integration with systems of record outside the platform.
• Define the security model, covering Dataverse security roles, business units, teams, column level security and hierarchy security, and align it to Entra ID groups.
• Set and enforce data loss prevention policies, connector governance and tenant level guardrails, typically through a Centre of Excellence pattern.
• Advise clients on licensing and capacity, including the cost implications of Premium connectors, Dataverse storage, Copilot Studio messages and Power Platform requests.
• Design, build and deploy custom agents in Copilot Studio for internal and client-facing use cases.
• Author topics and conversational flows, and decide deliberately where to use authored topics and where to rely on generative orchestration.
• Ground agents in enterprise knowledge, including SharePoint, Dataverse, public and internal websites, and file-based sources, and tune retrieval quality rather than accepting the default behaviour.
• Build actions that let agents do work rather than only answer questions, using connectors, Power Automate flows, agent flows, custom APIs and Model Context Protocol servers.
• Implement authentication correctly, including Entra ID single sign-on and user context propagation.
• Deploy agents across the right channels, most commonly Microsoft Teams, Microsoft 365 Copilot, SharePoint and web.
• Test, evaluate and improve agent behaviour over time, using analytics, conversation transcripts, escalation rates and topic-level performance to drive changes.
• Apply responsible AI practice, including scope limits, refusal behaviour, escalation to a human, and clear communication to users about what the agent can and cannot do.
• Extend agents into Microsoft 365 Copilot where appropriate, including declarative agents and connected knowledge.
• Build canvas applications with well-structured Power Fx, reusable components and a considered approach to delegation and performance against large data sources.
• Build model-driven applications on Dataverse, including forms, views, business process flows and command bar customisation.
• Develop custom components with the Power Apps Component Framework where the out-of-the-box controls are not sufficient.
• Design interfaces that people will actually use, including responsive layouts, accessibility and offline behaviour where the scenario calls for it.
• Build cloud flows for approvals, integration, scheduled processing and event-driven automation.
• Build desktop flows for robotic process automation against systems that offer no API.
• Develop custom connectors for internal and third-party APIs.
• Apply error handling, retry policy, concurrency control and monitoring.
• Integrate solutions with SharePoint Online, Teams, Exchange and OneDrive.
• Use Microsoft Graph where the platform connectors fall short.
• Integrate with Azure services such as Azure Functions, API Management, Logic Apps, Service Bus and Azure OpenAI where a requirement exceeds what the platform provides natively.
• Manage solutions properly, including managed and unmanaged solution strategy, solution layering and environment variables.
• Build deployment pipelines using Power Platform pipelines or Azure DevOps, with source control and automated deployment.
• Maintain documentation that allows a client team to take ownership after handover.
• Own the technical direction of an engagement.
• Review the work of other developers and set standards for the team.
• Act as the primary technical point of contact for the client.

Benefits

• Experience designing and delivering training, whether formal client enablement sessions, internal upskilling, or structured handover to a client team at the end of an engagement.
• Relevant Microsoft certifications, for example the Power Platform Solution Architect, Power Platform Developer or Power Platform Functional Consultant tracks.
• Experience establishing or running a Power Platform Centre of Excellence.
• Experience with Power Pages for external-facing portals.
• Experience in a consulting or client-facing delivery environment.
• Familiarity with the Model Context Protocol and with agent patterns outside the Microsoft stack.
• Traditional development background, for example C# or TypeScript, for cases where the platform is not the right answer.