2025’s Gartner® CFO & Finance Executive Conference featured sessions and hands-on product demonstrations from finance and operations providers covering growth strategies, cost management tools, and of course, AI in finance. I’ve had the pleasure of attending for a few years now, and what struck me this year was the excitement and enthusiasm for not just learning about but applying agents and generative AI to financial processes. Based on the conversations I had with customers at the event, here are the three things finance leaders should know about getting started with agents and AI.
Automate high-impact processes first
The journey to AI-powered finance doesn’t require a complete overhaul—it can start with targeted, high-impact steps that build momentum and deliver measurable value. Start with areas that are ripe for automation and deliver immediate impact to your team and bottom line. Financial close, account reconciliation, and supplier communications are ideal entry points using prebuilt agents available from Microsoft, like the Supplier Communications Agent in Microsoft Dynamics 365 Supply Chain Management, which can automatically respond to vendor emails, assess supply chain disruptions, and update purchase orders—helping to ensure continuity without manual intervention.
The breadth of agent possibility doesn’t stop with what’s available today as a prebuilt solution. Microsoft empowers finance professionals to build custom agents for ERP processes, applying automation and intelligence to critical processes, like the process at the heart of finance: record to report. Imagine the impact agents can have helping your team update tax strategies in a complex macroeconomic environment or surfacing real-time financial audits. Microsoft Copilot Studio agent builder provides a simple interface to quickly and easily build agents like the Financial Insights Agent, a tool designed internally for Microsoft to help finance professionals quickly surface and analyze key financial metrics. By using Azure OpenAI and semantic search, this agent allows users to ask natural language questions like “What’s my total operational expenditure (OPEX) as a percentage of expenses?” and receive instant, contextual answers. It can significantly reduce time spent on manual data retrieval and analysis while enhancing decision making with real-time insights. This agent exemplifies how AI can transform planning and analysis workflows, helping make finance teams more agile and strategic. Build your own agent here.
Establish a Center of Innovation to upskill your finance team
Create a dedicated team or initiative to explore, test, and scale AI use cases across finance. This “Center of Innovation” can serve as a hub for experimentation, governance, and best practices—helping ensure that AI adoption is intentional, secure, and aligned with business goals. Creating space for experimentation and testing means giving finance teams the freedom to test new tools, workflows, and ideas without the pressure of immediate perfection. It involves setting up low-risk environments, like sandboxes or pilot programs, where teams can trial AI agents, automate processes, and iterate quickly. This approach fosters a culture of curiosity and continuous improvement, allowing teams to scale what works and learn from what doesn’t.
Establishing a Center of Innovation was central to the growth and success of Microsoft’s own AI adoption. Microsoft Director of Modern Finance, Cory Hrncirik, joined the Microsoft session to share tangible examples of both AI adoption and the structures put in place to get us there.
As team members begin to shift from transactional to strategic roles, it’s imperative to equip your team with the skills required to work alongside AI. This could include training on prompt engineering for your function, data literacy, and how to manage and monitor AI agents.
Microsoft Copilot Studio offers a low-code environment where finance professionals can build and customize agents without needing deep technical expertise.
Work with IT to define your human-agent operating models to scale what works
As agents take on more operational tasks, CFOs must define how humans and AI will collaborate. This includes setting clear roles, responsibilities, and oversight mechanisms. Think of your team as a hybrid workforce—where agents handle the repetitive and analytical, and humans focus on judgment, strategy, and innovation. Due to the sensitive nature of financial material, early collaboration on security and permissions with IT will be crucial to implementation and driving outcomes.
Getting started with Microsoft and Microsoft Dynamics 365
If you’re ready to start reimagining ERP processes with agents, learn more about the pre-built agents announced in April 2025, explore how the new Model Context Protocol servers for Dynamics 365 will accelerate the ability for customers to build agents, and visit “AI-Powered ERP Solutions” for more product information.
Want to dig deeper? Learn how your organization can benefit from using Dynamics 365 Finance and Dynamics 365 Supply Chain Management in a three-year projection window1:
- 106% return on investment (ROI).
- USD8.9 million in productivity gains from unified data access, streamlined processes, and automated workflows.
- USD1.2 million in profitability improvements from real-time visibility and enhanced decision making.
- USD3.9 million in infrastructure savings by moving from legacy on-premises systems to the cloud.
These benefits are grounded in real-world outcomes and are explored deeper in a Forrester Total Economic Impact study, commissioned by Microsoft. The study’s composite organization is modeled after interviews with Dynamics 365 customers, and the report explores how teams reported significant improvements in operations.
Source:
1The Total Economic Impact™ Of Microsoft Dynamics 365 ERP, Cost Savings And Business Benefits Enabled By Dynamics 365 ERP, a Forrester Total Economic Impact™ Study commissioned By Microsoft, April 2024
GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved.
发表回复