Role-Based Programme
RB0572

Microsoft Copilot Studio for Finance & Accounting

Build Intelligent Agents for Finance Operations, Controls & Workflow Automation

IMAGE REQUIRED
Duration
16 Hours
Level
Intermediate
Delivery
Instructor-Led
Format
Capability Training

Programme Objectives

  • Develop functional proficiency in designing Microsoft Copilot Studio agents for finance and accounting processes.
  • Build knowledge-grounded and system-connected agents for finance enquiries, transaction support, approvals, reconciliations, close activities, and reporting workflows.
  • Integrate generative orchestration with deterministic Agent Flows, connectors, business rules, and human approvals.
  • Develop reusable finance workflows for exception management, document processing, status tracking, notifications, and operational coordination.
  • Apply testing, monitoring, security, segregation-of-duties, auditability, and governance controls to finance-agent implementations.

Tools covered

Microsoft Copilot StudioAI AgentsGenerative OrchestrationKnowledge SourcesCopilot ConnectorsPower Platform ConnectorsAgent FlowsHuman-in-the-LoopDataverse IntegrationScheduled & Event-Based TriggersAgent EvaluationMonitor & AnalyticsSecurity & Governance Controls

Who should attend

  • Finance Managers
  • Accounting Managers
  • Financial Controllers
  • Finance Operations Professionals
  • Accounts Payable Professionals
  • Accounts Receivable Professionals
  • General Ledger Professionals
  • Record-to-Report Professionals
  • Procure-to-Pay Finance Professionals
  • Order-to-Cash Professionals
  • FP&A Professionals
  • Finance Analysts
  • Shared Services Finance Professionals
  • Finance Transformation Professionals
  • Finance Process Owners

Prerequisites & Participant Readiness

  • Working knowledge of finance or accounting processes
  • Familiarity with invoices, expenses, journals, reconciliations, approvals, reporting, or financial controls
  • Basic awareness of ERP, finance, or business applications is recommended
  • General understanding of workflow automation and AI concepts is helpful
  • No programming expertise required
  • No previous Microsoft Copilot Studio course completion required

TOC Modules

Concepts
  • Understanding agents, knowledge, tools, connectors, triggers, and Agent Flows
  • Understanding generative orchestration and multi-step agent planning
  • Differentiating conversational assistance, deterministic workflow automation, and autonomous actions
  • Identifying appropriate finance use cases across AP, AR, GL, FP&A, shared services, and financial operations
  • Understanding why financial approval and accounting judgement require defined human accountability
Practical activities
  • Exploring a representative finance agent and its architecture
  • Mapping Finance Process → Agent Role → System → Human Control → Business Outcome

Scenarios

Supplier Invoice Exception to Controlled Resolution

Supplier Invoice Query → Copilot Studio Agent → ERP / Finance Record Lookup → PO & Invoice Validation → Exception Identification → Agent Flow → Finance / Business Approval → System Update → Supplier / Requester Status

Participants build an intermediate finance workflow that combines natural-language interaction, transactional data, deterministic rules, exception management, and authorised human approval without allowing the AI agent to make independent payment decisions.

Month-End Close to Finance Management Review

Close Calendar → Scheduled Agent Flow → Task Status Collection → Outstanding Reconciliations → Exception Identification → Responsible Owner Follow-Up → Validated Financial Information → Management Summary → Controller Review

Participants design a coordinated month-end process that automates routine tracking and communication while keeping reconciliations, accounting adjustments, financial interpretation, and final close approval under qualified finance ownership.

## Current Capability Reference

Microsoft currently defines **generative orchestration** as an LLM-driven planning layer that interprets intent, breaks complex requests into steps, selects relevant tools and knowledge, and executes multistep plans with safety and compliance guardrails. ([Microsoft Learn][1])

Current **Agent Flows** provide deterministic automation and can be triggered manually, by agents, by events, or on schedules. They support AI actions, branching and looping controls, data operations, Microsoft and third-party connectors, and **Human-in-the-Loop** activities such as approvals or requests for additional information. ([Microsoft Learn][2])

Current Copilot Studio **Agent Evaluation** supports repeatable test sets that can be rerun after changes to instructions, knowledge, tools, or workflows. Results can be compared across runs, helping finance teams validate whether changes improve or degrade agent behaviour before production use. ([Microsoft Learn][3])

Current monitoring provides visibility into agent activity, including usage, session details, tools and knowledge sources used, failures, and other performance information. Microsoft also provides enterprise controls covering **data policies, connectors, authentication, knowledge sources, publishing, environment governance, auditability, and security risk management**. ([Microsoft Learn][4])

Related programmes

Continue with programmes from the same capability area.

Take the next step

Ready to make this programme work for your team?

Customise modules, duration and business scenarios for your team.

Instructor-ledVirtualHybrid

Designed around your roles, tools and real workflows.